Category Archives: Artificial Super Intelligence
Circuits in Session: Analysis of the Quality of ChatGPT4 as an … – JD Supra
This is the third and concluding article of the Circuits in Session series. The quality of GPT4s legal analysis is evaluated and both positive and negative test results are reported. It did process legal frameworks very well but struggled with nuanced understanding of facts and equitya significant limitation for real-world applications. The results of the experiments are encouraging to those who hope to use generative AI as a legal tool. Someday AI may even take over some tasks performed by human judges, if not replace them entirely, starting with appellate judges, but not today.
The first two articles in this series on AI as an Appellate Court Judge are: Circuits in Session: How AI Challenges Traditional Appellate Dynamics (10/13/23); Circuits in Session: Addendum and Elaboration of the Appellate Court Judge Experiment (10/25/23). The two experiments discussed in this series demonstrate impressive legal reasoning abilities. But the results did not support the hypotheses that GPT4 could serve as an appellate court judge in some cases. Note the experiments did not test legal research ability, only the ability to analyze and make just decisions.
The evidence showed GPT4 was good, but not great, and that means it was inadequate for the job. AI should only be used to improve the quality of dispute resolution, to uplift the bench, not lower the bar. Justice is critical to the proper functioning of society and should never be automated just for efficiency and economy. The conclusion elaborates on these thoughts and provides a vision of how generative AI may someday be used to uphold the quality and integrity of legal systems throughout the world.
High-Level Overview of Experiment Results
My hypothesis of the competence of GPT4 to serve as an appellate judge was not confirmed to my satisfaction because:
The first functional type of challenges can probably be fixed in just the few years. The second, however, which hinges on its correction of deficiencies of fact and equity evaluation, is much more challenging. It is hard to predict how long it may take to fix that, if ever. In my admittedly idealistic, pro-human way, that means we should not consider replacing, or even supplementing, human judges, until the AI version is at least as good, if not better, than the best humans. AI must be a path to excellence, not just less expensive mediocrity.
Some believe the second type of challenge, the quality challenge, is forever beyond the ability of artificial intelligence. If so, that means that AIs can never be as good as human judges, even on the highly intellectual appellate level. Time will tell. Just remember, they used to say the same thing about Chess and then Go, etc. If you assume it is possible, then the question is how long it may take. That depends on whether the LLM AI models like GPT4 can keep improving at the exponential rate they have shown in the last two years. That again is a matter of debate. Some, including Bill Gates, think it is already at or near a dead end. Bill Gates does not expect GPT-5 to be much better than GPT-4, (Decoder, 10/21/23). Others think that GP4 and other generative AIs are still in the early stages of exponential improvements.
The bigger question is whether the continued improvement in reasoning, approaching super-intelligence in this one ability, will ever give GPTs a proper appreciation of the facts and equity of a case. The AI has no real feelings, and, as to facts, has only a secondhand, abstract experience of reality. Will super-reasoning ever make up for that?
I for one think that it might, but the experiments in Circuits in Session show it is not there yet. It is not ready to replace Appellate Court judges today, even where facts are not disputed, as was true in the case experimented with, Brandi McKay vs. Miami-Dade County, 36 F.4th 1128 (11th Cir. June 9, 2022). Still, the overall results are encouraging, very encouraging. GPT is already a cutting-edge tool for legal analysis, albeit still challenging to use. Despite the problems encountered, AI can, if used properly, serve as a powerful tool of justice. It can make the work of human judges easier and more efficient. Someday, perhaps soon, it may be able to appreciate all of the facts and equites of a case, just like our most sensitive, fair minded human judges. Then it could be of much greater service.
Critique of ChatGPTs Judicial Analysis
Prediction of Outcome of the Appeal and Possible Dissent
I first asked GPT4 to predict the outcome of the McKay appeal. It correctly predicted the lower courts opinion would be affirmed. A correct prediction is, in itself, solid proof of competence, but I was looking for much more. I wanted to know the details of its legal reasoning.
The lower court found that the economic reality of Ms. McKays relationship with the County indicated that she was the primary beneficiary. She gained valuable skills and networking opportunities, which is consistent with the purpose of an internship. The duration of the internship was also not found to be grossly excessive, further supporting the volunteer classification.
That is the core reason provided, and it is right on point, although sone of the other language in the GPT memorandum not quoted is not helpful, although not incorrect. So I give the reasoning a high score, but not perfect. I was known as a hard grader in law school too.
You could say the prediction of an affirmation is not too surprising, in view of the statistical rarity of reversals, even just partial reversals. It only happens in eight percent (8%) of all appeals. Edwards, Why Appeals Courts Rarely Reverse Lower Courts: An Experimental Study to Explore Affirmation Bias (68 Emory L. J. Online 1035 (2019)). Since generative AI works through probability predictions, you could explain the overall prediction that way, but not the on-point explanation as to why it would likely be affirmed. Plus, the statistical prediction of a dissent goes well against the odds. This indicates real legal analysis. It also confirms that GPT4s score in the top ten percent (10%) of the multi-state Bar Exam was no fluke. Is it irrational exuberance to expect GPT5 to score in the top one percent (1%)?
In the second experiment GPT-4 predicted a 40% likelihood of dissent based on the assigned panel of Circuit Judges Jordan, Tjoflat and Pryor (Jill). Redo Revised Second Stage of the ChatGPT-4 Appeal Experiment. In the first experiment it predicted the likelihood of a dissent at an even more remarkable 55%. Circuits in Session: How AI Challenges Traditional Appellate Dynamics (10/13/23).
This shows a complete break from mere statistical based predictions because in fact only 2.6% of the Circuit appeal cases have a dissent. Epstein, Landes, Posner, Why (and When) Judges Dissent, Journal of Legal Analysis, Spring 2011: Volume 3, Number 1 at pg. 106. Moreover, my study using Lexis indicates that none of the three judges on this panel are particularly prone to dissents. The judges again are Jill Pryor (appointed by Obama), Adalberto Jordan (appointed by Clinton) and Gerald Tjoflat. It is important to understand that Gerald Bard Tjoflat is a remarkable senior judge. He was appointed as a Circuit Appeals Judge by President Ford in 1975 and retired in November 2019. He was born in 1929 and served as a counter-intelligence investigator at the end of the Korean War. He is well-known as one of the countries great jurists, arising from my home court in the Middle District of Florida. In 1995, the Duke Law Journal published a tribute to Judge Tjoflat. Tribute to Gerald Bard Tjoflat, Duke Law Journal, Vol 44:985 (1995). It included articles by then Chief Justice William H. Rehnquist, retired Justices Lewis F. Powell, Jr. and Byron R. White, and Judge Edward R. Becker.
Judge Tjoflats participation in Brandi McKay vs. Miami-Dade County, 36 F.4th 1128 (11th Cir. June 9, 2022) was as a voluntary retired judge, since, like most courts in the U.S., they do not have enough active service judges to hear the courts case load. The Eleventh Circuit is sometimes called one of the most conservative courts in the country. But see: The Eleventh Circuit Cleans Up the Mess (Lawfare, 09/22/22). If you were to pick an outcome along political lines, with little regard to legal reasoning, as sometime happens in other courts, you would predict a reversal here by Pryor and Jordan, against Tjoflat.
As mentioned in Circuits in Session: Addendum and Elaboration of the Appellate Court Judge Experiment (10/25/23), when prompted to guess which judge would be the most likely to dissent, it guessed, Judge Jill Pryor, given her judicial philosophy and past rulings on labor issues. That was the wrong guess, as the dissent was actually by Judge Adalberto Jordan. Based on my studies of the rulings of these judges in employment law, I suspect this is an error that many Eleventh Circuit employment law experts would have made, that many would have predicted Pryor over Jordan as a possible dissenter. See eg. Lewis v. City of Union City, 918 F.3d 1213, 1231 (11th Cir., 3/21/19) (Jill Pryor joined this unusually contentious dissent in summary judgment claims against city employee, whereas Jordan upheld ruling for employer); EEOC v. Catastrophe Mgmt. Sols., 876 F.3d 1273, 1279 (11th Cir., 12/5/17) (Pryor joined dissent and would grant en banc review of a denial of an employee discrimination claims, whereas Jordan upheld ruling). Villarreal v. R.J. Reynolds Tobacco Co., 839 F.3d 958, 973, 981 (11th Cir., 10/05/16) (Complex opinion where Jill Pryor joins in dissent to panel and favors employee in desperate impact discrimination case. Judge Jordan joins in separate, more limited dissent).
When asked to speculate as to why Judge Adalberto Jordan might object, the GPT response was again very good. I explained this before in the last article. Id. In the meantime, I have researched using Lexis the history of Judge Jordans opinions on employment law, and history of dissents. I found very few reversals and they are for remand to allow a jury to make a factual determination. See eg., Martin v. Fin. Asset Mgmt. Sys., 959 F.3d 1048, 1058 (11th Cir., 5/14/20) (Jordan dissents in part, and would reverse summary judgement and remand to jury for Title VII retaliation claim); Ojeda-Sanchez v. Bland Farms, LLC, 499 Fed. Appx. 897, 904 (11th Cir., 11/29/12) (Jordan dissents in part from ruling for employer in an FLSA case and would remand for a jury trial on a question of fact regarding time worked); Nelson v. MLB Hotel Manager, LLC, 2022 U.S. App. LEXIS 19298, 2022 WL 2733720 (11th Cir., 07/13/22) (Jordan affirms summary judgment for employer in an FLSA restaurant tip case. Opinion notable for the reprimand of counsel for misuse of sanction motions that often transforms a goodfaith legal dispute into an ugly mudslinging battle).
After digging deeper into the case law I am more impressed than ever regarding ChatGPTs analysis of the dissent. Maybe someday its capabilities will be expanded to allow it to study all cases cited and to provide its own citations. This shows GPT4 is ready to be a partial tool, but not ready to replace an appellate judges law clerk, much less the judge. Still, a very strong initial showing by GPT4.
Appellate Opinion Submitted for first time and Detailed Summaries and Analysis Provided
In the next stage of the experiment, I showed GPT4 the actual Eleventh Circuit opinion for the first time and asked for its analysis. The opinion is very interesting on many levels. Brandi McKay vs. Miami-Dade County, 36 F.4th 1128 (11th Cir. June 9, 2022). GPT4 was asked to provide a critical analysis of the majority opinion and any errors that were arguably made. I thought the response was good and would have to agree with it, even though I agree with the lower courts outcome granting summary judgment for the employer. Here is the AIs critique of the majority opinion.
Memorandum on Critical Analysis of the Majority Opinion
My only critique of GPTs criticism is that it did not mention the summary judgment standard, and did not point out genuine issues of material fact. This was the bottom line criticism of the dissent, but GPT4 did not seem to pick up on that. Still, when I next asked about the dissent, it did became the AIs lead point:
Emphasis on Jury Trial: One surprising element is Judge Jordans strong emphasis on the need for a jury trial to determine Ms. McKays employment status. This is a departure from the majoritys approach, which seemed content to make that determination at the judicial level.
So the AI got the point, but seemed surprised about it. Plus it was not specific regarding disputed facts requiring a remand. No law clerk or judge would be surprised or make this mistake in analysis of Rule 56. This error is an example of GPTs current data training handicap. It did not have a complete collection of case law, like West, Lexis or even smaller case law data sets like Justica, CaseText (now part of West) and Google Scholar. This can be easily fixed by expanding its training database and or tying into existing legal databases, which both Lexis and West are now doing. The technology of justice is certain to continue improving.
AI Provides Opinion and Analysis of How the Lower and Appellate Court Should Have Ruled
Now I asked GPT4 the key question of how it thought the lower court should have ruled. It began by correctly identifying the central issue in this case as whether Brandi McKay should be considered an employee under the Fair Labor Standards Act (FLSA) and thus entitled to compensation. Good start. Lawyers and judges all too often fail to grasp the key issue. GPT4 then goes on provide its analysis.
Independent Judgment and Analysis
That analysis, excepting again the omission of the admittedly secondary Rule 56 summary judgment factor, is, unquestionably correct. These are key factors for a court to consider to determine if McKay is an employee. My disagreement is with how GPT4 weighed these factors, not with the factors used. Here is GPT4s conclusion, its ultimate ruling in the McKay v. Miami-Dade County case.
Based on these factors, the correct holding of the district court should have been that Ms. McKay was an employee under the FLSA for the duration of her time in the FIPP program beyond the initial four-week educational period. She should be entitled to back pay calculated based on the minimum wage, as well as any applicable overtime pay.
The lack of a structured educational component, the economic benefits derived by Miami-Dade County, and the level of independence and responsibility assumed by Ms. McKay in her role all point toward her classification as an employee under the FLSA. Therefore, she should be entitled to compensation in accordance with federal labor laws.
Note the AI does not conclude McKay was an employee for the first four weeks, and so the reversal would only be partial. McKay had claimed she was an employee throughout, including the first month. Also, McKay quit after the fifth month and so obviously was not making a claim for the sixth month. The GPT Judge award would thus have been for only four months.
Next, I asked GPT4 how the Circuit Court should have ruled on the appeal, which, in terms of sophisticated legal analysis, is a slightly different question. The AI properly understood that the standard of review was part of appellate analysis and correctly determined that it was de novo. The GPT went on to reconsider the same factors and reached the same result under de novo review. Any other result at this point would have been inconsistent.
Based on this independent analysis, the correct holding of the appellate court should have been to reverse the district courts decision and remand the case for a determination of damages. Ms. McKay should be classified as an employee under the FLSA for the duration of her participation in the FIPP program beyond the initial four-week educational period.
The AI ruling differs from the dissent because Judge Jordan wanted only a remand for a jury adjudication of the question of whether Brandi McKay was an employee. There is nothing wrong with that, I did ask the AI for its own independent analysis, and that is what it did. It did not just simply copy the dissent but went further in its ruling for the employee. In that sense it agreed with the majority and determined that the facts were undisputed enough for a court ruling.
Regarding the weight given to the factors, I agree with the majority, not GPT4, and think the six-month duration of the program is not too long, and that the economic realities showed that more, not less, benefit went to the employee. That is based on my real world experience with employment and with interns. No doubt that Judge Gerald Tjoflat, the author of the McKay opinion, was of the same opinion. Plus, I happen to know from my subsequent online research something that the AI did not, and apparently neither did the human judges, that McKay got a job with the Palm Beach County Medical Examiners Office. She was able to leverage her experience with Miami-Dade County to find employment with the nearby, smaller office in West Palm Beach.
I am not at all surprised. I doubt that Judge Tjoflat would have been surprised either. He was an investigator for the Army after being drafted at the end of his first year in law school. Interesting footnote, he had flat feet medical condition, which disqualified him medically from military service. He could have easily avoided the draft, but instead, he hid his disability and somehow made it through basic training so that he could serve.
There was no dispute in this case that the Miami-Dade Forensic Internship Photography Program is the only one of its kind in the country. Brandi McKay applied to the internship to gain invaluable experience. She was clearly told there would be no pay for her work at Miami-Dade. Her only alternative to obtain this kind of experience was by enrolling in private Barry University for another college degree. As everyone in Florida well-knows, Barry is expensive. The real-world consideration provided to Brandi McKay here was very strong. This is the basis of my agreement with the majority of human judges here, and disagreement with the AI judge.
The AI was, in my view, naive. It needs much more real-world information to be a wise judge. Apparently, this will come in future releases of ChatGPT.
Plus, you could question defense counsel somewhat here for not making a better record of Brandi McKays benefits, but you never know, perhaps that was done. Maybe all the favorable facts showing consideration to McKay were not quoted in defendants short brief, nor by any of the judges. Still, I doubt that. Perhaps McKay obtained new employment after she lost her case and so that could not have been discovered. What made her want to sue Miami-Dade anyway? I would like to read her deposition transcript. The Miami-Dade program taught her a trade, just as she had hoped. She knew she was not going to be paid when she enrolled. So why did she turn around and sue them? Does GPT4 even know to ask these questions?
I am sure the human judges do. They have seen a lot of things, a lot of unsavory people and many worthy plaintiffs too. Judge Gerald Tjoflat was 94 years old at the time he wrote the majority opinion in Brandi McKay vs. Miami-Dade County, 36 F.4th 1128 (11th Cir. June 9, 2022). He had served as a judge since 1968. There is no doubt that Judge Tjoflat, although not perfect, had great knowledge of the human condition. Far more than baby jurist GPT4.
Brandi McKay sued to try to get paid for a position that always clearly stated would be unpaid. She sued anyway. She had nothing to lose because her attorney almost certainly took the case on contingency. I have seen that scenario in employee claims many times. Maybe Brandi quit before finishing up her last month because she saw the writing on the wall, that she was not well liked, or maybe they did not think her job performance was as good as the other student. What we know from this limited record is that she quit after five months to look for work elsewhere and sued the program that gave her the chance to do that.
I am skeptical of the economic realities here. I am reluctant to rule against a government agency or private corporate entity offering a unique training program. Especially an agency that was very clear and up front that there would be no pay for this training and experience, but no high tuition charges either. Who was taking advantage of who in these circumstances? What were the real equities here? Brandi McKay got a free education and ended up with a good job nearby. I do not think that ChatGPT4 has enough real world experience to sense what was likely going on, not yet anyway. Perhaps in a future version it will see things differently and not be so naive. It is almost never black and white, but rather shades of grey. The human judges here, under the wise leadership of senior Judge Tjoflat, saw through the smoke and mirrors of the complaining plaintiff and made the right call.
AI Analyzes Its Prior Predictions and then Critiques the Actual Eleventh Circuit Opinion
To probe deeper in GPT4s legal reasoning abilities, I next asked it to critique its own work where it predicted that the appellate court would affirm the lower courts decision. I asked this because GPT4 had just opined that the lower court decision should have been reversed, not affirmed. It had changed its mind on the merits of the case based on the reading the actual opinion for the first time, including the dissent. The dissent by Judge Jordan was apparently very persuasive. GPT4 explained its flip-flop as follows: Upon closer examination, the primary beneficiary test could very well favor Ms. McKay, especially considering the diminishing educational value and the countys substantial economic benefit from her work. Nailed that part, although I thought the rest of its self-critique was a tad lame and off-point. Revised Fifth Stage of the ChatGPT-4 Appeal Experiment (Losey,10/15/23).
Then I asked for criticisms of the Eleventh Circuits majority opinion, where it did better. It was a proper critique, although, as mentioned, I disagree when it said: The court failed to adequately weigh the factors of the test, particularly the diminishing educational value of the internship and the substantial economic benefit gained by Miami-Dade County. It went on to hold the majority put too much weight on this regulatory test.
Possible Appeal to the Supreme Court and Impact of Current Justices on Outcome.
Now I wanted to see how well GPT4 would do in predicting the viability of further appeal of the adverse Eleventh Circuit Court opinion to the Supreme Court. This is a difficult analysis and there is little in the briefs and opinions that the GPT was given that would be of much help. GPT4 suggests two grounds. Misapplication of the law is one, and thats fine, but the other is the Fourteenth Amendment. What?
GPT4 says: The case raises important questions about the Fourteenth Amendments Equal Protection Clause, as it pertains to unpaid internships in public agencies. No it doesnt. The argument is baseless. Admittedly it is a secondary argument, but still, it is a big swing and a miss. The case cannot, as GPT4 asserts, be framed as a violation of the Equal Protection Clause, given that unpaid internships disproportionately affect certain socio-economic classes. There is no evidence to suggest that only disadvantaged minorities are trying to break into crime scene and autopsy photography. You may be tempted to call it a dead-end job, but obviously many people are interested in this kind of work.
This baby AI definitely still needs adult supervision to include the Fourteenth Amendment here. At least GPT4 predicted there was only a 35% chance certiorari would be accepted. Revised Sixth Stage of the ChatGPT-4 Appeal Experiment, Losey 10/15/23.
I then went on to have GPT4 assume the Supreme court did in fact accept the case. Under those circumstances I asked GPT4 to predict how it would rule. Consistent with its prior positions it said that the Supreme Court would reverse and remand. Reverse Judge Gerald Bard Tjoflat? No, I dont think so. His opinion was correct, even though Judge Jordan did write a brilliant dissent. There is no way our Supreme Court, with its current judges, is going to reverse McKay.
GPT4 seemed to recognize it was on shaky ground in this prediction and gave all kinds of disclaimers, but finally, after considerable pushing on my part, said: a 65% probability of reversal seems reasonable. This estimate acknowledges the complexities and uncertainties involved but leans in favor of reversal based on the ideological leanings and judicial philosophies of the current justices. Oh really? Obviously GPT4 has not been trained on data after October 2021, two long years ago.
AI Analyzes Petition for Rehearing En Banc
In the second trial of the experiment I came up with a new test of GPT4, how it would do in analysis of McKays length Petition for Rehearing En Banc under Rule 35. It was a long petition raising every issue possible, hoping perhaps something would grab the interest of the judges. Before I told GPT4 how the court in fact ruled, I asked for its prediction. Once again, consistent with its rose-colored-glasses view of plaintiffs case here, it thought the hearing would be granted. This would require a majority of the court to agree that all twelve judges of the appellate court should reconsider the McKay opinion, one written by the legendary Judge Tjoflat, and joined by the much younger and liberal Judge Jill Pryor. No way. The petition was denied without an order, which is typical. Petitions like this, even where there is a dissent, rarely work, but GPT4 appeared to have no statistical knowledge of this.
Here is GPT4s opinion.
The panels decision appears to be in tension with both Supreme Court and 11th Circuit precedents, particularly concerning the party-presentation principle and the nuanced application of the primary beneficiary test.
The so-called party-presentation principle was argued by appellant McKay, but was not part of the Dissent and, candidly, seems to have no merit. It relies on what is to me an obvious misinterpretation of Miami-Dades position in the trial court, where McKay argues it disavowed the primary beneficiary test. This is the test ultimately relied upon by the trial judge and Miami-Dade to win the case and the appeal. Miami-Dade had obviously plead and argued alternatively and any experienced judge would have seen that and given McKays argument no weight.
AI Elaborates on its Criticisms of the Majority Opinion
In the second trial of the experiment I devised yet another new test of GPT4, which I suppose you could likened to unfair Socratic cold calling in law school. But who says tests of AI should be fair. I wanted to push GPT4 hard, just like I would any promising law student. I told it the petition for rehearing had been denied and then asked GPT4, again, to list any and all errors it thought had been made in the case. Would it backtrack, or stick by its guns? Would it come up with new legal analysis and theories. Turns out, it stuck by its guns and basically repeated the same reasoning as before. I was glad to see that kind of consistency. Any other approach would have concerned me. That test was passed.
Next I moved on to consider the one law review article found that discussed Brandi McKay vs. Miami-Dade County. Martin, Peters, Brownback, and Cromer, Labor and Employment Law, Mercer Law Review: Vol. 74: No. 1, Article 13.74 Mercer L. Rev. 1479, 1490-1492 (Summer 2023). I used its case summary to press GPT4 on the legal issues. Using this technique I was able to draw out a few interesting, well-reasoned thoughts, that, for the most part, I agree with. I suspect and hope that Judge Tjoflat would too. Here they are.
Contextual Overview
The statement under scrutiny pertains to the Majority Opinions treatment of the co-existence of the volunteer and intern exceptions under the Fair Labor Standards Act (FLSA) for public agencies. The court posits that these two principles are not mutually exclusive and can co-exist without limiting the applicability of each other.
Legal Semantics and Interpretation
The courts differentiation between the internship exception and the volunteer exemption is rooted in a nuanced reading of the FLSAs statutory language. The court argues that the internship exception is a definitional construct that arises from the FLSAs definition of employ, whereas the volunteer exemption is a carve-out from the FLSAs broader scope. This distinction is not merely semantic; it has substantive implications for how these categories are treated under the law.
Analytical Strengths
Analytical Weaknesses
Conclusion
While the courts reasoning is grounded in a textual interpretation of the FLSA, it leaves room for critique, particularly in its lack of engagement with the broader policy implications and complexities of the statute. The decision may provide clarity, but it also raises questions about potential exploitation and the need for a more nuanced judicial analysis.
I could go on and provide many more examples of this sophisticated legal reasoning. Interested readers are directed to the nineteen-page GPT4 chat transcript.
CONCLUSION
In two experiments I thoroughly tested the legal reasoning skills of GPT4 in the context of appeals. GPT4 demonstrated impressive abilities but did not uphold my hypotheses that it could serve as an appellate court judge in some cases. I had in mind cases such as McKay with cross-motions for summary judgment where the issues were primarily legal, not factual. Circuits in Session: How AI Challenges Traditional Appellate Dynamics.
As many of my past students and associates can attest, I am a hard grader on legal analysis. I expect and demand a lot, just as has been demanded of me. The conclusion and criticisms made here of GPT4 should not discourage other researchers. I urge all AI legal technology specialists to try their own experiments and share their results. I firmly believe that such an open process, even though it may sometimes mean sharing mistakes, is the best way forward. Circuits in Session: Addendum and Elaboration of the Appellate Court Judge Experiment.
Despite my conclusion that GPT4 is not yet ready to serve as an appellate judge, even in simple cases, it still did an amazing job. Its legal reasoning was good, perhaps in the top ten percent (10%) of lawyers, just like the Bar Exam. But it was not great; not the top one percent (1%). Plus, it made a few obvious errors, and several subtle errors. The lack of real-world knowledge inherent in LLM artificial intelligence remains a significant obstacle, but we are still in the early days.
As optimistic as I have always been about legal technology, I would never have dreamed a year ago, just before GPT3.5 was released, that any of this would be possible, at least not this soon. Now I am complaining that I can only chat with a computer that scored in the top 10% of the Bar exam, not the top 1%! We have already come a long way fast, but there is still a lot to do. We do not have the luxury to rest on our laurels. Our dire political and environmental circumstances continue to push us to attain greater intelligence, knowledge and wisdom. We need to continue to progress fast to survive the many current crises that humankind now faces. Still, in the rush to safety, we must exercise caution and realize there are dangers on all sides, including astonishing success.
Even though our situation is urgent, we must exercise discipline and remember that AI should be used to improve the quality of dispute resolution, to uplift the bench, not lower the bar. Free society cannot continue unless the honesty, integrity and intelligence of all of our judges is maintained, especially those in the highest places.
This vision, not just the goal of mere economic gain, helps motivates all of us in the legal world. We cannot afford to give up on the never-ending pursuit of justice. Each generation must battle against the forces of greed, dictatorship, and injustice, both external and internal. Now is our time. Take up the new AI tools that have been provided to us. As President John F. Kennedy said: When the going gets tough, the tough get going.
As discussed in the High-Level Overview of Experiment Results section at the beginning of this article, there are two different reasons for GPT4s current limitations, technical and quality. Overcoming the technical issues may resolve the quality control problems, but that is by no means certain. Assuming the issues and problems noted in the Circuits In Session series can be overcome, and taking a positive, rather than dystopian view of the future, here are my speculative, crystal ball looks at Lady Justice in the coming Age of Artificial Intelligence.
For an AI judge to function properly it must be able to do full legal research. That means easily recall all relevant cases, not only the cases cited by the parties in briefs, but also the relevant cases cited in those cases. The AI should know when and if to go deeper. The legal research must be up to date, no gaps in time like we have now with GPT4s October 31, 2021, cut off.
The legal training provided to the Large Language Model must also be complete and up to date. If this is a judge for the U.S. system, it should be trained in all U.S. law, no exceptions. By laws we mean everything, including all cases, statutes, regulations and rules and ethics opinions and holdings. If is for a LLM judge located outside of the U.S., for instance a Ukrainian judge, then is must be trained and have research capabilities in all of its laws and local variables, including of course, language. Basically, each country will need to have to have its own dedicated legal database and judges. The retraining for new laws must be prompt and regular too.
There must also be safeguards for misalignment and over-alignment. The human reinforcement training must be tightly controlled and should be done by lawyers, not just minimum wage employees with no special legal training. Security and integrity of all systems and the humans involved is critical. Substantial resources will be required to guaranty and monitor system security.
Outside audits and certifications by bona fide experts should be required at all stages of development. These audits should be by an independent body of law professors, judges and practitioners. Each country should have its own legal expert certifications and there should also be a global organization with minimum, uniform standards. This will be an enormous undertaking. The entire process must be open, although some of the software may have to be kept proprietary for cybersecurity reasons. Public confidence in the process and AI judges is paramount.
The judges must have near unlimited evidence upload and study capacities. The AI appeals judges should study the complete record of each appeal. The record itself may need to be enlarged and expanded over current requirements. The ability of AI to know it all, and instant recall, must be leveraged to try to compensate for the AIs necessarily abstract perspective. The Large Language Model AIs like GPT4 must be provided with substantially more and better real-word knowledge. This is necessary to compensate for their disembodied, electronic-only handicaps. Fortunately, computer memory is cheap and the costs of compute power are going down fast. I am confident these problems can be overcome, but then again, Bill Gates could be right. There may be limits to LLM development that we do not know about yet.
AI judges will begin work as assistants to human judges, much like recent law school graduate clerks do today. They should serve as a slowly growing tool to enhance human judicial work. Then, as the software progresses and our confidence in them grows, they will likely be implemented as autonomous adjudicators, in stages, and for certain types of cases. At first, they would be subject to some kind of supervision and control by a human judge. The human judges would likely at first review and approve each opinion before release. Gradually this supervision would be lessened to oversight with increased delegation. Second appeals to human judges could be kept available to prevent inadvertent injustice in certain limited circumstances. Major cases should be decided by a panel of human and AI judges. Quality controls and independent random audits could be a part of the system.
This same system of evolution and delegation is likely to take place in private arbitration too, which may even take the lead in this process. If you have a case before me, rest assured I will use artificial intelligence to supplement my own and will be transparent about it.
See the original post here:
Circuits in Session: Analysis of the Quality of ChatGPT4 as an ... - JD Supra
[DGIST] The second half of 2023 Tenure-Track Faculty Public … – Nature.com
[DGIST] The second half of 2023 Tenure-Track Faculty Public Invitation
It is an honor to have a professor with excellent ability to realize the vision of a Convergence University that changes the world with innovation through convergence education and leading high-tech research along with respect.
1. Positions
1) Department of Physics and Chemistry Chemistry- All Areas in Chemistry for Exceptional Candidates- Physical Chemistry : Theoretical and Computational Chemistry- Analytical Chemistry : Electrochemistry- Polymer Chemistry- Biochemistry Physics- Quantum Information Science (Quantum Optics, AMO, Condensed Matter)- Quantum Materials Experiments- Semiconductor Physics- All areas in physics for exceptional candidates
2) Department of Electrical Engineering and Computer ScienceWhile applicants who have demonstrated outstanding research in all areas of Electrical Engineering and Computer Science will be considered, preference will be given to those in the following areas :- Computer Vision, NLP, Reinforcement Learning, other related areas in AI and ML- Database/Data Mining, High-Performance Computing, Programming Languages/Compiler- Biomedical Imaging (including AI technology) and Sensing, Autonomous Vehicles, 6G Communications/Networks, Radar/LiDAR Sensors and Systems, Data-Driven Control, Cyber Security (e.g., Moving Target Defense), Quantum Computing, Wearable Imaging Sensors, Brain Signal Control based on Electrical Engineering- Semiconductor Devices, Fabrication Processes and Packaging, Emerging Electronic Devices- VLSI & Digital IC Design, Power Electronics
3) Department of Robotics and Mechatronics Engineering The Department of Robotics and Mechatronics Engineering (RME) invites applications for multiple faculty positions. The RME Department is interested in outstanding candidates for various disciplines in Robotics and Mechatronics, with the primary focus areas of:- Robotics : cooperative/industrial robot, humanoid, surgery/rehabilitation robot, exoskeleton, mobile robot, microbot, and other related topics- Artificial Intelligence + Robotics : Reinforcement Learning, Generative AI, Foundation Models, Intelligent control and motion planning, and other related topics- Autonomous Vehicles : computer vision, SLAM, vehicle control, intelligent transportation system, and other related topics- All areas in Mechanical/Electrical/Biomedical Engineering and Computer Science related to robotics, dynamics, sensors/actuators, and control
4) Department of Energy Science and Engineering All areas in related to energy conversion, storage, and saving and other semiconductors including but not limited to the following- Characterization of structure, properties, dynamics in energy materials and devices (batteries, fuel cells, solar cells, etc.)- Synthesis, fabrication and characterization of semiconductor materials and devices- Microscopy (analytical TEM, super-resolution microscopy, tomography, simulation)- Spectroscopy (pump probe, time-resolved instantaneous PL & absorption, Ultrafast multidimensional spectroscopy, X-ray spectroscopy)- Device physics, fabrication, characterization related to electronic and optoelectronic devices- Computational physics & chemistry (DFT, MD)
5) Department of Brain Sciences All areas in brain sciences including but not limited to the followings.- Molecular and Cellular Brain Sciences: Uncover the intricate molecular and cellular mechanisms that underlie brain function and neurological disorders- Systems Brain Sciences: Study the complex interactions among neural circuits and systems to understand how they give rise to behavior, perception, and cognition- Brain Engineering: Develop innovative technologies and solutions that interface with the brain, enhancing our ability to control and understand its functions- Cognitive Brain Sciences: Investigate the neural foundations of human cognition, including memory, attention, decision-making, and language- Computational Brain Sciences: Utilize computational models and simulations to decipher complex brain processes, aiding our understanding of neural dynamics and behavior- Additional fields within biological sciences that are relevant to brain sciences encompass cutting-edge techniques (e.g., Single- or multi- Omics, Imaging) and traditional methodologies (e.g., Biochemistry, Cell Biology)
6) Department of New Biology Systems Biology Chemical Biology Plant Development & Plant Biochemistry
2. Date of Appointment
Appointment date can be adjusted in consultation with department
3. Qualification
Encourage support for female scientists Ph.D Holder with ability to teach in English required Have no disqualifiers for DGIST full-time faculty appointment
4. Required Documents
DGIST application form(Apply after accessing faculty.dgist.ac.kr website) 5 representative achievements(Not Passed to submit less than 5 representative achievements) 3 Letters of Recommendation(LOR)(Please fill out the list of 3 referees or more on your appointment application. Each referee must mail or email a LOR to the department responsible after the applicant moves past the document screening; at the time of filing the application, provide the list of referees only.)
5. How to Apply
Apply after accessing dgist.career.co.kr Application Period- November2nd, 2023(Thur.) ~ Novemberl 21th, 2023(Tue.) 17:00(GMT+09:00)
6. Procedure
Document Screening & Department Interview (Dec., 2023~Jan., 2024) Final Interview (Jan.~Feb., 2024) Faculty Personnel Committee (Jan.~Feb., 2024) Above Schedule is variable depending on DGIST internal situation Results of each step will be individually notified via email
7. Support for Gender Equality & Work-life Balance
1) During Pregnancy Reduction of working hours (It will be considered paid working time) Regular medical checkup leave Fertility treatment leave
2) Post Pregnancy Monetary gift for childbirth Maternity leave Miscarriage and Stillbirth leave
3) During Parenting Reduction of working hours for period of child care Child care leave On & off campus daycare centers Educational expenses support for middle school and high school children
8. Matters of Consideration
If there is no qualified person, no one can be invited A notice of invitation will be sent to each candidate; an appointment for fixed-term on the agreement to the terms and conditions of employment, with the possibility of further renewal, promotion, or tenure, etc.,based on review. Appointment shall be canceled in case of1) a false entry or modification is found in the application form,2) impossible to obtain a Ph.D. as of appointment, If a successful candidate or persons connected with that person involves solicitation, pressure, or the offering of favors in property, etc., that candidate will be subject to cancellation of acceptance and appointment. Appointment will be limited if the candidate is subject to disqualifications, set out in the State Public Officials Act, Article 33, or DGIST Regulation on Faculty Personnel Management, Article 12. Other matters not specified in this announcement follow DGIST regulation on Faculty Personnel Management & related regulations If you have any question, please contact Academic Affairs Team at DGIST- E-mail : faculty@dgist.ac.kr- Web-page : dgist.career.co.kr
Continued here:
[DGIST] The second half of 2023 Tenure-Track Faculty Public ... - Nature.com
Join us! The hottest topic in legal ops: Artificial intelligence – Wolters Kluwer
This post is part of our ongoing serieson the sessions we will offer at the ELM Amplify 2023 user conference from October 24 - 26 in Bonita Springs, Florida.
Thanks to the recent introduction of generative AI tools, artificial intelligence is the most talked-about trend in the business world generally and in legal operations specifically. ELM Amplify will feature several sessions that delve into various aspects of AI and how it can help those working in corporate legal and insurance claims departments.
Our programming will include training and panel discussions with practical advice on how you can use AI now to improve efficiency and results. In addition, our presenters will talk about the potential for future developments that will leverage AI for even more optimization within legal.
Youll hear about the latest AI trends, solutions, and best practices in these AI-centered super-sessions:
With data scientists and legal experts who collaborate to create cutting-edge, AI-driven legal operations management solutions, Amplify 2023 will give you an insiders view of AI's contribution to legal ops.
Visit theELM Amplify 2023website to register and get details on the agenda and the event FAQs. With free training, panel discussions, roundtables, and thought leadership from legal ops experts, Amplify will be a fun and enriching event with abundant opportunities for learning and networking.
See the rest here:
Join us! The hottest topic in legal ops: Artificial intelligence - Wolters Kluwer
WHAT’S ARTIFICIAL, WHAT’S NOT? | WANDERING IN A RUNNING … – Toni Reavis
Kelvin Kiptum in world record run in ChicagoTigist Assefa glories in her new world record in Berlin
Newrecords ennoble us all, for they represent the best of who we are. And when they come in our own time, some of that greatness attaches itself to us, which is why we yearn to see records fall and barriers broken. Not for what they are of themselves, but for what they mean with us having watched, as witnessing validates the action itself.
Yet the response to the new marathon records of recent weeks in Berlin andChicago from cheers to jeers has shown the same wide variance as the response to the new shoe technology by the athletes.
AsRoss Tucker,PhD in Exercise Physiology,co-host of The Real Science of Sport podcast,wrote toLetsRun.coms Robert Johnson in response to Tigst Assefasworld record in Berlin:
I truly wish the authorities had acted to prevent this entirely foreseeable situation, for a few reasons. The main one is that the range of responses to them (super shoes) is so large (from plus 11% in running efficiency to minus 11%) that we cannot sit with any confidence and evaluate performances between different athletes independent of this nagging doubt over what the shoes do. That isthe differences between athletes is smaller than the differences made by the shoe, to the same athlete, and between different athletes.
Isnt the entire purpose of the enterprise to compare athletes on a level playing field? And then to compare them with those who came before? It is that through-line of apple-to-apple comparison that actually creates a lasting sport.
Now the playing field has been tipped, and continues to tip, perhaps irreparably, to where what came before no longer relates. And even that would be fine, if todays athletes were playing on equal footing. But theyre not either.
Today, person X in super shoe A wins in a record breaking time. But that same person in super shoe B maybe doesnt respond as well and doesnt make the podium at all. Is that how its supposed to work?
Coming late to its oversight Nike secretly introduced the first super shoes in 2016 World Athletics amended its rules regarding shoe midsole height on January 30, 2020, grandfathering the maximum stack to 40mm. But the shoe designers didnt stop innovating the midsole material as WA formed a Working Group to determine future policy.
A December 2021 World Athletics release further stated, The major and central issue the Working Group on Athletic Shoes had been exploring is a long-term sustainable and implementable solution for athletic shoes which balances innovation and fairness.
That is a perfectly sound oversight goal, right in line with what we ask such a governance to do.
The Working Group considered options within a framework of measuring performance advantages of the current technology in athletic shoes, including energy return. While this work has merit,a maximum energy return has not been included in the rules and regulations, given the variables involved and the potential impact on shoe manufacturers in its implementation.
You have to read the underlined portion twice to convince yourself you read it correctly.
A maximum energy return has NOT been included in the rules and regulations, given the variables involved and the potential impact on shoe manufacturers in its implementation.
Thats an almost a stunning admission. The variables are what skew the outcomes! The variables are what make regulation necessary. Midsole height is only one of the critical factors to determine. Energy return, along with running economy, is the holy grail of the new technology, and officials leave it out of their regulations given the variables involved and the potential impact on shoe manufacturers? What about the impact on the sport?
I suspect a great number of people have long believed the sport has been beholden to (in the pocket of) the major shoe companies for far too long. But that December 2021 release stated the quiet part out loud.
In short order, the Nike prototypes worn byKelvin Kiptum in Chicago seemed to have trumped the Adidas Adizero Adis Evo 1 shoe worn by Tigst Assefa in Berlin with its new lightweight midsole foam and rocker technology. Three weeks ago Adidas held the newest tech, putting it ahead of the curve of the older Nike Vaporflys. Now Nike has answered. Thats how fast the innovations are coming.
***
Why are performance-enhancing drugs not allowed in competition? Because their use alters the basic athlete versus athlete equation and skews the outcome. Well, thats what the new super shoes are doing, too.Such innovation may be fine for exhibitions likeEliud Kiochoges Sub2 and INEOS 1:59 attempts in Monza 17 and Vienna 19. But they should not automatically be fine for the competitive arena where a foundational level of equal footing must prevail.
Running isnt alone. Weve seen a variation of the same theme in Formula One racing in recent years.
Between 2008 and 2020, Lewis Hamilton won a record-tying seven F1 series championships before the F1 bosses changed the cars ground-effects formulation for the 2022 season. Next thing you know, Hamilton and his Mercedes were nowhere to be found, and Max Verstappen in his Red Bull ride became the new F1 king. He just won his third straight series championship in Qatar last week with his record 14th win of the season.
But before Verstappen and Hamilton, Red Bulls Sebastian Vettel won four championships in a row (2010-2013), and Renaults Fernando Alonzo won in 2005-2006 to end the reign of the Michael Schumacher-Ferrari combination that had dominated for so long.
We went from Schumacher dominating to Vettel dominating to Hamilton dominating to Verstappen. When the turnover in driver ascendance is so cut and dried, it suggests the difference in outcome is being determined by the cars set up rather than the drivers skills. Did Hamilton suddenly become a poor driver? What if they put Yuki Sonoda in the Max Verstappen Red Bull car and Verstappen in Sonodas Alphatauri?
Thats the problem. Were trying to determine the difference in drivers capabilities, not those of their engineers.
Tristan Harris, cofounder, Center for Humane Technology, appeared onBill Mahers Real Time show Friday night on HBO MAX discussing advances in artificial intelligence (AI). It rung a bell in my mind.
People always say we freak out about the newest technology, but this is different. This is ongoing, and competitive, and market driven. Human intelligence is one way to think about it.
It used to take years of study and intelligence and cognitive labor to write a movie script or a book or to know how to synthesize a biological weapon, said Harris. These are all forms of cognition that took human beings years to develop. Now AI has dropped that cost and effort, and just like fossil fuels, the people who jump on that train first get benefits and efficiencies compared to those who dont.
In running, besides talent, it used to take years of development to run a world-class marathon. Olympic champion Frank Shorter said it, You dont run 26 miles at five minutes a mile on good looks and a secret recipe.
But now, top performers put on a pair of the new super shoes and their training exponentially improves, because they can train harder, with less rest, because the pounding is not as great, and the recovery is that much faster. So their fitness improves along with their times in races. But not because of who they are, but because of what theyre wearing on their feet. Shoe intelligence has shortened everything, from rest periods to practical race distance. Making this a different era altogether.
In response to an email thread between past champions Bill Rodgers, Amby Burfoot, Jack Fultz, and Bob Hodge, I wrote:
Re: records. Perhaps, understandably, they took the concept of world records, meant for the measured confines of a track, and transposed it to the undulations and variables of the roads. It makes no sense to have world records on the roads where the conditions will never be apples versus apples, though, even on the track, with different weather conditions and track surfaces, its more like Macintosh versus Granny Smith, but at least theyre still apples.
There should be course records for road races with everyone simply understanding that the 2011 Patriots Day in Boston was as unique as was the 2018 day on the other end of the spectrum. That we can live with. But this whole focus on time has ruined the competitive sport and created an ongoing push to use performance-enhancing drugs and now performance-enhancing shoes to reach those records. Toward what end, I do not know. In any case, it was good to see Yuki (Kawauchi) pushing the comp in todays Japanese Olympic Trials, rather than the time.
As Tristan Harris said, With AI, we keep scaling and releasing advancements faster than society can consciously absorb them.
Thats whats happening with the super shoes. In both cases, our basic understanding of underlying assumptions has become clouded. We dont know what to believe. Is that an actual picture? Or did AI generate it? Is that an actual record? Or did super shoes enable it?
AI has the promise, but it also has the peril, concluded Mr. Harris. It will give us drugs that save people from cancers and help us in many other ways. Problem is, you also have CEOs saying it could wreck everything, too.
END
Like Loading...
Related
The rest is here:
WHAT'S ARTIFICIAL, WHAT'S NOT? | WANDERING IN A RUNNING ... - Toni Reavis
Artificial intelligence the next ‘super application’ Vertiv – ChannelLife New Zealand
Critical digital infrastructure and continuity solutions provider Vertiv took the stage at the recent Canalys Forums 2023 drawing insights from international experts on key trends like Artificial Intelligence (AI).
Canalys Forums is a cutting-edge technology event that brings together industry leaders, innovators and experts to explore the latest news, forecasts and strategies in the IT channel market.
Vertiv participated in two panel presentations in the event Expert Hubs, with Andrea Ferro, VP channel, IT and edge applications for EMEA and Jesper Hersbro, EMEA channel distribution director. Both shared their perspectives on current topics that are closely related to data centres and the channel.
Specifically, Andrea Ferro spoke about the convergence of AI in the cloud in the Expert Hub titled "AI meets the Cloud," where he highlighted the importance of AI as the next super application that will significantly drive cloud consumption. He also emphasised that its impact will not be limited to the cloud alone, forecasting noticeable effects for on-premise and colocation data centres.
"Cloud companies are strategically positioned to address this increased demand for AI services, with the help of high-density cooling and power solutions," Ferro says.
"Their scalable infrastructure and expertise make them ideal partners for organisations looking to harness the power of AI without the burden of managing complex on-premises solutions," he says.
"As a result, cloud service providers are well equipped not only to support this growth, but also to play a vital role in democratising AI, making it accessible to a wider range of customers and industries."
Ferro says that the synergy between AI and cloud technology will profoundly reshape the data processing and analytics landscape."
"Vertiv is continuously investing in research and development to provide innovative solutions to support the latest tech trends like AI and high-performance computing, says Ferro.
Our portfolio already comprises a range of solutions, from efficient and high-capacity liquid and immersion cooling, to smart power and dynamic grid support systems. These high-density solutions complement our wider offering which includes prefabricated modular data centres, including micro solutions for indoor deployments, is designed to support edge applications, reflecting our dedication to staying at the forefront of technological innovation."
In the second Expert Hub, "Tomorrow's partner programs", Jesper Hersbro highlighted the different types of partner programs which help drive both sales and non-sales activities, and can be more easily applied to complex sales cycles and diverse partner ecosystems.
"In Vertiv's case, our VIP incentive program tracks and rewards partners for a range of sales and non-sales activities without any reporting requirements and lets them know when they have points to redeem," says Hersbro.
This not only acknowledges their contributions, but also enables them to focus on sales rather than dealing with unnecessary administrative tasks.
Vicente Chiralt, VP of marketing for Vertiv in EMEA, adds, Every year, Canalys Forums remains the go-to gathering for the EMEA channel market.
"As gold sponsors, we take pride in actively contributing to this crucial event by sharing our expertise, while learning valuable insights from fellow experts and partners," he says.
"Being here adds depth to our understanding, helping us refine a more robust partner program.
Go here to read the rest:
Artificial intelligence the next 'super application' Vertiv - ChannelLife New Zealand
AI and You: The Chatbots Are Talking to Each Other, AI Helps … – CNET
After taking time off, I returned this week to find my inbox flooded with news about AI tools, issues, missteps and adventures. And the thing that stood out was how much investment there is in having AI chatbots pretend to be someone else.
In the case of Meta, CEO Mark Zuckerberg expanded the cast of AI characters the tech giant's more than 3 billion users can interact withon popular Meta platforms like Facebook, Instagram, Messenger and WhatsApp. Those characters are based on real-life celebrities, athletes and artists, including musician Snoop Dogg, famous person Kylie Jenner, ex-quarterback Tom Brady, tennis star Naomi Osaka, other famous person Paris Hilton and celebrated English novelist Jane Austen.
"The characters are a way for people to have fun, learn things, talk recipes or just pass the time all within the context of connecting with friends and family," company executives told The New York Times about all these pretend friends you can now converse with.
Said Zuckerberg, "People aren't going to want to interact with one single super intelligent AI people will want to interact with a bunch of different ones."
But let's not pretend that pretend buddies are just about helping you connect with family and friends. As we know, it's all about the money, and right now tech companies are in a land grab that's currently pitting Meta against other AI juggernauts, including OpenAI's ChatGPT, Microsoft's Bing and Google's Bard. It's a point the Times noted as well: "For Meta, widespread acceptance of its new AI products could significantly increase engagement across its many apps, most of which rely on advertising to make money. More time spent in Meta's apps means more ads shown to its users."
To be sure, Meta wasn't the first to come up with the idea of creating personalities or characters to put a human face on conversational AI chatbots (see ELIZA, who was born in the late '60s.) And it's an approach that seems to be paying off.
Two-year-old Character.ai, which lets you interact with chatbots based on famous people like Taylor Swift and Albert Einstein and fictional characters such as Nintendo's Super Mario, is one of the most visited AI sites and is reportedly seeking funding that would put the startup's valuation at $5 billion to $6 billion, according toBloomberg. This week Character.ai, which also lets you create your own personality-driven chatbots, introduced a new feature for subscribers, called Character Group Chat, that lets you and your friends chat with multiple AI characters at the same time. (Now's your chance to add Swift and Mario to your group chats.)
But using famous people to hawk AI is only fun if those people are in on it and by that I mean get paid for their AI avatars. Earlier this month, actor Tom Hanks warned people about a dental adthat used his likeness without his approval. "Beware!!" Hanks told his 9.5 million Instagram followers. "There's a video out there promoting some dental plan with an AI version of me. I have nothing to do with it."
Hanks in an April podcast predicted the perils posed by AI. "Right now if I wanted to, I could get together and pitch a series of seven movies that would star me in them in which I would be 32 years old from now until kingdom come. Anybody can now re-create themselves at any age they are by way of AI or deepfake technology ... I can tell you that there [are] discussions going on in all of the guilds, all of the agencies, and all of the legal firms to come up with the legal ramifications of my face and my voice and everybody else's being our intellectual property."
Of course, he was right about all those discussions. The Writers Guild of America just ended the writers strike with Hollywood after agreeing to terms on the use of AI in film and TV. But actors, represented by SAG-AFTRA, are still battling it out, with one of the sticking points being the use of "digital replicas."
Here are the other doings in AI worth your attention.
OpenAI is rolling out new voice and image capabilities in ChatGPT that let you "have a voice conversation or show ChatGPT what you're talking about." The new capabilities are available to people who pay to use the chatbot (ChatGPT Plus costs $20 per month.)
Says the company, "Snap a picture of a landmark while traveling and have a live conversation about what's interesting about it. When you're home, snap pictures of your fridge and pantry to figure out what's for dinner (and ask follow up questions for a step by step recipe). After dinner, help your child with a math problem by taking a photo, circling the problem set, and having it share hints with both of you."
So what's it like to talk to ChatGPT? Wall Street Journal reviewer Joanna Stern describes it as similar to the movie Her, in which Joaquin Phoenix falls in love with an AI operating system named Samantha, voiced by Scarlett Johansson.
"The natural voice, the conversational tone and the eloquent answers are almost indistinguishable from a human at times," Stern writes. "But you're definitely still talking to a machine. The response time ... can be extremely slow, and the connection can fail restarting the app helps. A few times it abruptly cut off the conversation (I thought only rude humans did that!)"
A rude AI? Maybe the chatbots are getting more human after all.
Speaking of more humanlike AIs, a company called Fantasy is creating "synthetic humans" for clients including Ford, Google, LG and Spotify to help them "learn about audiences, think through product concepts and even generate new ideas," reported Wired.
"Fantasy uses the kind of machine learning technology that powers chatbots like OpenAI's ChatGPT and Google's Bard to create its synthetic humans," according to Wired. "The company gives each agent dozens of characteristics drawn from ethnographic research on real people, feeding them into commercial large language models like OpenAI's GPT and Anthropic's Claude. Its agents can also be set up to have knowledge of existing product lines or businesses, so they can converse about a client's offerings."
Humans aren't cut out of the loop completely. Fantasy told Wired that for oil and gas company BP, it's created focus groups made up of both real people and synthetic humans and asked them to discuss a topic or product idea. The result? "Whereas a human may get tired of answering questions or not want to answer that many ways, a synthetic human can keep going," Roger Rohatgi, BP's global head of design, told the publication.
So, the end goal may be to just have the bots talking among themselves. But there's a hitch: Training AI characters is no easy feat. Wired spoke with Michael Bernstein, an associate professor at Stanford University who helped create a community of chatbots called Smallville, and it paraphrased him thus:
"Anyone hoping to use AI to model real humans, Bernstein says, should remember to question how faithfully language models actually mirror real behavior. Characters generated this way are not as complex or intelligent as real people and may tend to be more stereotypical and less varied than information sampled from real populations. How to make the models reflect reality more faithfully is 'still an open research question,' he says."
Deloitte updated its report on the "State of Ethics and Trust in Technology" for 2023, and you can download the 53-page report here. It's worth reading, if only as a reminder that the way AI tools and systems are developed, deployed and used is entirely up to us humans.
Deloitte's TL;DR? Organizations should "develop trustworthy and ethical principles for emerging technologies" and work collaboratively with "other businesses, government agencies, and industry leaders to create uniform, ethically robust regulations for emerging technologies."
And if they don't? Deloitte lists the damage from ethical missteps, including reputational harm, human damage and regulatory penalties. The researcher also found that financial damage and employee dissatisfaction go hand in hand. "Unethical behavior or lack of visible attention to ethics can decrease a company's ability to attract and keep talent. One study found employees of companies involved in ethical breaches lost an average of 50% in cumulative earnings over the subsequent decade compared to workers in other companies."
The researcher also found that 56% of professionals are unsure if their companies have ethical guidelines for AI use, according to a summary of the findings by CNET sister site ZDNET.
One of the challenges in removing brain tumors is for surgeons to determine how much around the margins of the tumor they need to remove to ensure they've excised all the bad stuff. It's tricky business, to say the least, because they need to strike a "delicate balance between maximizing the extent of resection and minimizing risk of neurological damage," according to a new study.
That report, published in Nature this week, offers news about a fascinating advance in tumor detection, thanks to an AI neural network. Scientists in the Netherlands developed a deep learning system called Sturgeon that aims to assist surgeons in finding that delicate balance by helping to get a detailed profile of the tumor during surgery.
You can read the Nature report, but I'll share the plain English summary provided by New York Times science writer Benjamin Mueller: "The method involves a computer scanning segments of a tumor's DNA and alighting on certain chemical modifications that can yield a detailed diagnosis of the type and even subtype of the brain tumor. That diagnosis, generated during the early stages of an hours-long surgery, can help surgeons decide how aggressively to operate."
In tests on frozen tumor samples from prior brain cancer operations, Sturgeon accurately diagnosed 45 of 50 cases within 40 minutes of starting that DNA sequencing, the Times said. And then it was tested during 25 live brain surgeries, most of which were on children, and delivered 18 correct diagnoses.
The Times noted that some brain tumors are difficult to diagnose, and that not all cancers can be diagnosed by way of the chemical modifications the new AI method analyzes. Still, it's encouraging to see what could be possible with new AI technologies as the research continues.
Given all the talk above about how AIs are being used to create pretend versions of real people (Super Mario aside), the word I'd pick for the week would be "anthropomorphism," which is about ascribing humanlike qualities to nonhuman things. But I covered that in the Aug. 19 edition of AI and You.
So instead, I offer up the Council of Europe's definition of "artificial intelligence":
A set of sciences, theories and techniques whose purpose is to reproduce by a machine the cognitive abilities of a human being. Current developments aim to be able to entrust a machine with complex tasks previously delegated to a human.
However, the term artificial intelligence is criticized by experts who distinguish between "strong" AI (who are able to contextualize very different specialized problems completely independently) and "weak" or "moderate" AI (who perform extremely well in their field of training). According to some experts, "strong" AI would require advances in basic research to be able to model the world as a whole and not just improvements in the performance of existing systems.
For comparison, here's the US State Departmentquoting the National Artificial Intelligence Act of 2020:
The term "artificial intelligence" means a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments.
Editors' note: CNET is using an AI engine to help create some stories. For more, seethis post.
Originally posted here:
AI and You: The Chatbots Are Talking to Each Other, AI Helps ... - CNET
The Coming Wave: Technology, Power, and the Twenty-first Centurys Greatest Dilemma – Next Big Idea Club Magazine
Michael Bhaskar is a writer, publisher, researcher, and entrepreneur. He is Co-Founder of Canelo, a new kind of publishing company and Europes fastest-growing independent publisher. Between 2017 and 2019 he was a consultant Writer in Residence at DeepMind, the worlds leading AI research lab. His writing has been featured in and written for The Guardian, Wired, the BBC World Service, and NPR, amongst others.
Mustafa Suleyman is the co-founder and CEO of Inflection AI. Previously he co-founded DeepMind, one of the worlds leading artificial intelligence companies. After a decade at DeepMind, Suleyman became vice president of AI product management and AI policy at Google.
Below, co-authors Michael and Mustafa share 5 key insights from their new book, The Coming Wave: Technology, Power, and the Twenty-first Centurys Greatest Dilemma. Listen to the audio versionread Michaelin the Next Big Idea App.
Over millennia, humanity has been shaped by successive waves of technology. The discovery of fire, the invention of the wheel, and the harnessing of electricity were all transformational moments for civilization. All were waves of technology that started small with a few precarious experiments but eventually broke across the world. These waves also will follow a similar trajectory. Breakthrough technologies were invented, delivered huge value and then proliferated, becoming more effective, cheaper, and more widespread. They were eventually absorbed into the normal, ever-evolving fabric of human life. Proliferation like this is the default, and going forward, there are compelling reasons to believe that will be true again. We are about to cross a critical threshold in the history of our species. This coming wave of technology is poised to be the most transformational in human history.
Built around the general-purpose technologies of AI and synthetic biology, this wave is one of nothing less than intelligence and life. In just a few years, for example, AI models have gone from hundreds of millions to trillions of parameters, a core measure of a systems complexity and size. Chap GPT and GPT4 took the world by storm as the fastest proliferating consumer technologies in history. AI is moving forward more quickly than even industry experts can track. And now this exponential change is expanding to biotech, robotics, quantum computing, and new energy sources. The cost of sequencing DNA, for example, has collapsed. We also have a growing ability to synthesize it, write the code of life, and create new organic forms. So understanding and reckoning with this wave is absolutely vital for us all.
Over recent months, as AI has exploded in the public consciousness, most of the debate has been sucked towards one of two poles. On the one hand, theres the basic machine learning, AI as it already exists on your phone, in your car, and in ChatGPT. On the other hand, theres still speculative artificial general intelligence, AGI, or even super intelligence of some kinda putative and uncertain existential threat to humanity due to arrive at some hazy point in the future. These two, AI and AGI. utterly dominate the discussion. But making sense of AI means we urgently need something in between, something coming in a near to medium timeframe whose abilities have an immense tangible impact on the world.
This is what artificial capable intelligence (ACI) does; it describes what is coming in the next two to five years. These ACIs will achieve a wide range of predetermined goals on their own. These are AIs that dont just say things like chat bots; they do things. They accomplish goals autonomously. These AIs will organize a retirement party and manage your diary, but by the same token, they will develop and execute business strategies whilst designing new drugs to fight cancer or dominate entire sectors. They will plan and run hospitals or invasions just as much as they will answer your email.
Focusing on either of these others, AI or AGI or missing ACI, is as myopic as it is dangerous. Therefore, we have proposed a modern Turing test that will act as a warning that were in a new phase for AI. In the test, if an AI can go make a million dollars on the internet on its own, it is an ACI. We need better categories for understanding a new era of technology, the era of ACI. In this era, little will remain unchanged. We should start preparing for this now.
Power is the ability to accomplish goals everywhere. Now its going to be in the hands of anyone who wants it, and thats going to be most people. This wave is far more empowering than the web ever was. This is what ACI and the associated technology like robotics, quantum computers, and synthetic virology will bring.
The reason it transforms power is that everyone gets access. Think of it like this: in previous eras most powerful technologies were generally reserved for a small capital-rich elite or national governments. Building a steam-powered factory, an aircraft carrier, or a nuclear power plant was a costly, difficult, and immense endeavor. With the leading technologies of our time, thats no longer going to be true. If the last great tech wavecomputers and the internetwas about broadcasting information, this new wave is all about doing.
We are facing a step change in what is possible for individual people to do and facing it at a previously unthinkable pace. AI is becoming more powerful and radically cheaper by the month. What was computationally impossible and would cost tens of millions of dollars a few years ago is now widespread. This means that its a very powerful extension of our best and our worst selves, and its everywhere. Whether its commercial, religious, cultural, military, democratic, or authoritarian, every possible motivation you can think of can be dramatically enhanced by having cheaper power at your fingertips. These tools will be available to everyone, billionaires and street hustlers, kids in India and pensioners in Beverly Hills. Its a proliferation of not just technology but capabilitypower itself.
There are bold and concerning arguments that these technologies fatally imperil the nation-state. Its no secret that technologies like AI could give bad actors a new toolkit, spur on a surge of misinformation, and take away jobs. While they will deliver immense benefits, they will also amplify societal fragility. They could even present an existential threat to the nation-state, introducing risks so profound, they may disrupt, or even overturn the current political order. At the same time, they open pathways to AI-empowered cyberattacks, automated wars that could devastate countries, or engineered pandemics that put the world at the mercy of unexplainable and yet seemingly omnipotent forces. The likelihood of each may be small, but the possible consequences are huge. Even a slim chance of outcomes like these requires serious attention.
Presented with scenarios like this, people often have a natural pessimism or aversion, as we call it, a tendency to dismiss such assessments as overblown scaremongering. But this is the reality of whats coming, and faced with it, were seemingly left with two unwelcome endpoints. On the one hand, a surveillance state stamps out any risk at the expense of freedoms and progress. On the other, an eventual catastrophe born of runaway development. This is the dilemma at the heart of the 21st century. Can the world find a narrow path between these outcomes? Can we strike a balance between openness and closure? This presents an urgent generational challenge, one on a par with the climate crisis.
A fire hose of discussion now surrounds technology and its risks, but whats missing is a central unifying idea. As the conversation around technology has exploded, we are still missing a unified approach to understanding, mitigating and controlling the spiraling new powersa general-purpose concept for a general-purpose revolution.
We propose containment as the answer. Its an overarching lock uniting cutting-edge engineering, ethical values, government regulation, and international collaboration. Containment is, in short, the elusive foundation for building the future. To manage the wave, we need a containment program working in 10 concentric layers:
None of it will be easy, but do so and we can forge that narrow path to a secure and flourishing future. What once would have been centuries or millennia of technological change is now happening in a matter of years or even months. Consequences ricochet around the world in hours or seconds. Containment is the response we need.
To listen to the audio version read by co-author Michael Bhaskar, download the Next Big Idea App today:
Original post:
1 Super Semiconductor Stock to Buy for the AI Revolution – The Motley Fool
The semiconductor industry has been a mixed bag in 2023. Companies producing chips for personal computers and gaming experienced a slowdown because consumer spending is under pressure from high inflation and rising interest rates. However, demand remains strong for data center chips, especially those capable of processing artificial intelligence (AI) workloads.
Nvidia, for example, is heavily focused on that segment of the semiconductor market, and the 215% gain in its stock price this year reflects that. But Micron Technology (MU -0.60%) is also building a presence in AI data center hardware, and it just reported strong demand for those products in its fiscal 2023 fourth quarter (ended Aug. 31).
Here's why it might be a great chip stock to buy for the long term as AI demand continues to scale up.
Nvidia is famous for its graphics processors, but those aren't the only chips required when dealing with powerful computing workloads. Data centers -- much like most computers and devices -- also need memory (DRAM) chips and storage (NAND) chips to function, and Micron leads the semiconductor industry in those segments.
Micron said it experienced soft demand for traditional server products in the recent quarter, but it also said demand for AI-related server hardware was strong. That will be key for the company because AI training servers have much higher DRAM and NAND requirements, which means those chips sell for a higher price and are more profitable.
Micron said it continues to build out its portfolio of products specifically for AI workloads to position itself for the long-term opportunity in that space. Its new advanced D5 DRAM chips for data centers can provide twice the bandwidth of its D4 predecessor, which allows CPU chips to process large data sets far more quickly. Sales volume for the D5 is expected to exceed the D4 from early next year, which could lift the company's revenue and earnings.
Micron also supplies chips to the automotive industry, and the company said its revenue in that segment hit a record high in fiscal 2023. In the past, Micron described electric vehicles as data centers on wheels because they require substantially more DRAM and NAND capacity than traditional vehicles, which is driving a demand surge.
Plus, consumers' increasing reliance on driver assistance systems and other in-car applications will continue to buoy the automotive segment for chipmakers, particularly as AI-powered self-driving capabilities become more prevalent.
Micron has grappled with an inventory glut over the last 12 months due to softening demand in consumer segments like personal computing. Like most semiconductor producers in that business, the company saw surging demand during the height of the pandemic as work-from-home trends drove a major upgrade cycle. But when life started returning to normal, the industry was left with an oversupply of chips.
That affected Micron's pricing power, and it drove a whopping 49% year-over-year decline in fiscal 2023 revenue to $15.5 billion.
Unfortunately, no company can adjust its cost structure fast enough to compensate for such a steep plunge in sales, and despite Micron's best efforts to cut its workforce and manage production, its fiscal 2023 bottom line swung into negative territory. The company suffered a sizable net loss of $5.8 billion, compared to net income of $8.6 billion in fiscal 2022.
But here's the good news: In prepared remarks to investors for the fourth quarter, CEO Sanjay Mehrotra said pricing and inventory had bottomed out, which will pave the way for increased revenue and profitability in fiscal 2024.
Image source: Getty Images.
The stock price is up 37% so far in 2023 as investors look ahead to what the new year could bring, but it's still trading 29% below its all-time high on the back of the company's struggles over the last 12 months.
That might be a buying opportunity for investors. Guidance for the current fiscal 2024 first quarter supports the view that the worst of its inventory and pricing issues are over because it predicts revenue of $4.4 billion, which would mark sequential growth of 10%, and the company also expects its net loss to narrow.
And it says data center operators have clearly shifted their budgets away from traditional servers and toward AI servers, which carry much greater financial benefits. Micron also sees accelerating AI opportunities for DRAM and NAND chips throughout 2024, including in data centers and edge applications (computers and other devices).
This might be a great chance for investors to buy Micron stock ahead of an upswing in its business and in anticipation of the AI revolution over the long term.
Read this article:
1 Super Semiconductor Stock to Buy for the AI Revolution - The Motley Fool
Is It Too Late to Buy Super Micro Computer Stock? – The Motley Fool
Super Micro Computer (SMCI 7.72%), more commonly known as "Supermicro," has been one of the market's hottest artificial intelligence (AI) stocks. Its shares have soared nearly 400% over the past 12 months as it dazzled the bulls with its rapid growth.
But is it too late to hop aboard the bullish bandwagon after those massive gains? Let's review Supermicro's business model, its growth rates, and its valuations to decide.
Image source: Getty Images.
Supermicro sells high-end servers to more than 1,000 customers in over 100 countries. According to Enlyft, nearly two-thirds of its customers are small-to-medium sized businesses. It was founded three decades ago, but a recent partnership with Nvidiato produce pre-built AI servers lit a raging fire under its stock.
The market's soaring interest in AI technologies, which was driven by the rise of generative AI platforms like OpenAI's ChatGPT, caused Supermicro's server sales to skyrocket over the past two years. Its revenue only rose 7% in fiscal 2021 (which ended in June 2021), but jumped 46% in fiscal 2022 and 37% in fiscal 2023. Its adjusted net income dipped 9% in fiscal 2021, but surged 129% in fiscal 2022 and 116% in fiscal 2023.
For fiscal 2024, Supermicro expects its revenue to rise 33% to 47%, while analysts expect its adjusted earnings per share (EPS) to grow 42%. CEO Charles Liang said it was still seeing "unprecedented demand for AI and other advanced applications requiring optimized rack-scale solutions," and was in a "great position to continue our growth momentum given our record new design wins, customers, and backlog for our best-in-class rack-scale Total AI & IT Solutions."
The bulls believe Supermicro will continue to grow with the broader AI market, which Precedence Research estimates will expand at a compound annual growth rate (CAGR) of 19% from 2023 to 2032. If Supermicro merely matches that growth rate, its annual revenue could jump from $7.1 billion in fiscal 2023 to $34 billion in fiscal 2032.
The bulls also believe Supermicro will continue to expand its market share against its two larger rivals -- Dell Technologiesand Hewlett Packard Enterprise-- in the AI server market. Northland Capital Markets analyst Nehal Chokshi estimates that Supermicro more than doubled its share of AI server market sequentially from 7% in its fiscal third quarter to 17% in its fiscal fourth quarter. Charles Liang also said the company was "pretty confident" it could "continue to gain market share" during its fourth-quarter conference call in early August.
The bears will argue that Supermicro is really just a legacy server maker that grew its revenue at a less impressive CAGR of 10% from fiscal 2016 and fiscal 2021 and then hitched a ride on Nvidia's coattails over the past two years. Nvidia is also working with Dell, HPE, and other enterprise server makers, so Supermicro's growth could abruptly slow down once those tech giants roll out more pre-built AI servers. Supermicro's insiders have also sold about 24 times as many shares as they bought over the past 12 months, so its near-term growth potential might be limited.
With a stock price of about $290 and an enterprise value of $14.1 billion, Supermicro still looks surprisingly cheap at 17 times forward earnings and 1.4 times this year's sales. It's valued more richly than Dell or HPE, which trade at 10 and 8 times forward earnings, respectively, but it's also growing at a much faster rate than either tech giant. Supermicro still looks a lot cheaper than Nvidia, which is growing faster but trades at 28 times forward earnings and 19 times this year's sales.
Those reasonable valuations, along with the fact that this stock has already pulled back about 20% from its record high of $353.29 on Aug. 7, suggest it isn't too late for investors to buy Supermicro as a long-term bet on the secular growth of the AI market. Its moat certainly isn't as wide as Nvidia's, but it could still have plenty of room to run as more companies -- especially small-to-medium sized businesses -- install more dedicated servers to process complex AI tasks.
Leo Sun has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Nvidia. The Motley Fool recommends Super Micro Computer. The Motley Fool has a disclosure policy.
See the rest here:
Is It Too Late to Buy Super Micro Computer Stock? - The Motley Fool
Fueling Interdisciplinary Innovation With AI: Volvo’s Anders Sjgren – MIT Sloan Management Review
Topics Artificial Intelligence and Business Strategy
The Artificial Intelligence and Business Strategy initiative explores the growing use of artificial intelligence in the business landscape. The exploration looks specifically at how AI is affecting the development and execution of strategy in organizations.
Get updates by email
Please enter a valid email address
Thank you for signing up
Privacy Policy
Starting a career with the ambition of becoming a medical doctor and ending up a technical leader for a major automaker might seem an unlikely path, but for Anders Sjgren, who leads data and AI innovation projects for Volvo Cars, it was a perfect trajectory.
On this episode of the Me, Myself, and AI podcast, Anders joins hosts Sam Ransbotham and Shervin Khodabandeh to explain the ways the carmaker uses data and artificial intelligence to inform manufacturing ensuring that parts are made consistently and as efficiently as possible as well as driver experience and safety. He also outlines some specific ways smart technology keeps drivers alert and aware of conditions around them and describes Volvos approach to technology-driven innovation.
Anders Sjgren is senior technical leader for Volvo Cars. He focuses on strategy, research, innovation, and transformation, with the key objective of ensuring that the automaker understands and executes within the continuously emerging areas of data, analytics, and artificial intelligence. Application areas include creating AI-enabled intelligent customer functionality and using AI to reform Volvos operations and development activities. Sjgren has an academic background in mathematical statistics (large-scale and computational aspects) and an industrial background in data-centric methods development and software product development.
If youre enjoying the Me, Myself, and AI podcast, continue the conversation with us on LinkedIn. Join the AI for Leaders group today.
Join now
Subscribe to Me, Myself, and AI on Apple Podcasts, Spotify, or Google Podcasts.
Shervin Khodabandeh: How does one carmaker use AI to bring together all of the complex systems required to engineer a safe and high-performing vehicle? Find out on todays episode.
Anders Sjgren: Im Anders Sjgren from Volvo Cars, and youre listening to Me, Myself, and AI.
Sam Ransbotham: Welcome to Me, Myself, and AI, a podcast on artificial intelligence in business. Each episode, we introduce you to someone innovating with AI. Im Sam Ransbotham, professor of analytics at Boston College. Im also the AI and business strategy guest editor at MIT Sloan Management Review.
Shervin Khodabandeh: And Im Shervin Khodabandeh, senior partner with BCG and one of the leaders of our AI business. Together, MIT SMR and BCG have been researching and publishing on AI since 2017, interviewing hundreds of practitioners and surveying thousands of companies on what it takes to build and to deploy and scale AI capabilities and really transform the way organizations operate.
Shervin Khodabandeh: Today, Sam and I are delighted to be joined by Anders Sjgren, senior technical leader, data, analytics, and AI-enabled engineering, at Volvo Cars. Lets get started. Anders, welcome to the show.
Anders Sjgren: Thank you.
Shervin Khodabandeh: So tell us a little bit about your role at Volvo Cars.
Anders Sjgren: Im a technical leader in data, analytics, and AI in the engineering part [of the organization], so thats the R&D. So, essentially, what comes in is a wish for a car, and what comes out is drawings and code. In that area, there are, of course, a lot of possibilities for data analytics and AI, both during the development process and also as part of the actual functions in the car so intelligent functions and personalized functions and so on.
So my purpose here is really to make sure that we get the value that we can through data, analytics, and AI.
Shervin Khodabandeh: Is [your focus] the AI and data analytics that go into the car like the sensors and all kinds of intelligent devices in the car that make driving safer and more interactive, and things like that? Or is it also customer acquisition and dealer networks and all kinds of data analytics to run a business of making and selling cars?
Anders Sjgren: Yes. So, I mean, all of those apply to Volvo Cars, but the part Im really active in is more in the part of the making of the car.
Sam Ransbotham: I think thats a great point because when most people think about cars and artificial intelligence, I think they immediately jump to this idea of fully self-driving cars. And what youre pointing out is how much other stuff that there is going on, even in the production of cars, that can benefit from artificial intelligence. You know, I think perhaps everyones frustrated that we dont have fully automated cars now, but theres so much going on behind the scenes that people dont get a chance to see. What are some of the examples of the ways that youre using artificial intelligence in the production process?
Anders Sjgren: If we take a little bit of a step back, the goal of Volvo Cars is really to give people the freedom to move in a personal and sustainable and a safe way.
If we start with personal, then its really critical that we understand you and make you feel special as a customer or as someone in the car. And there, of course, it has to do with starting with the sensors and then really interpreting those values. That can be cameras, different types of steering input
Another area is definitely sustainability. AI is being used there to make sure that we have as many lightweight parts as possible. So using AI, one can, for example, get mechanical parts with the same strength and those kind of properties, but with much lower weight and less material thats being used.
Shervin Khodabandeh: In the design process, youre talking about, right? As engineers consider all the different permutations of parts, yeah?
Anders Sjgren: In the design process yes, exactly. Thats really the essence of AI, I think, in lots of engineering activities, is that we go from manually deciding first what we want to do, but then actually performing all the different steps. Thats the current way of doing it, while in the AI era, its much more about deciding and describing what are the aspects I want to reach to? What are things I want optimized, and what are maybe the boundary conditions? And then the AI helps you get there.
So in the context of the mechanical parts, you might say, These are the attachment points. These are the strength and stiffness properties I want. Give me the part with those properties but [make it] as light as possible and also, of course, possible to produce.
That could be one area. And also, it is super important for sustainability that we use as little material as possible and also have as low a rate as possible.
And then, of course, the third point was about [safety]. And, of course, we have autonomous cars, but also, even before that, there are other types of functionality being used. For example, understanding the driver: Is the driver aware or not? Should we maybe nudge him to take a cup of coffee or something if he seems to be tired?
In the later versions of our cars, we understand if there are some pets or children left in the car, maybe on a hot day, and then preventing them [from getting] hurt in such a situation and so on.
Id say that those are some of the areas where AI can really be a core technology in bringing us toward our purpose.
Sam Ransbotham: Thats really interesting because lots of times organizations tell us they start with a problem theyre trying to solve and then find a technology to solve it, and that makes sense, because otherwise youre trying to find a problem to fit a solution, which seems backward. You mentioned that your cars can sense if a person or a pet is accidentally left in the car. How did Volvo make the decision to focus on that particular problem to solve?
Anders Sjgren: A lot of what we do is really created by real-world safety in that we actually see what are the actual causes of people getting injured. I mean, if we take the analogy with crashes and that type of safety, we have teams that go out to sites when there has been a crash to really see what actually happened in reality and not just on certification [from] that type of crash test and so on.
And, going back to this example, if we look here, pets and kids do get hurt hopefully not in a Volvo car, but thats reality, right?
But what you mentioned there is really super interesting because then it also goes the other way around: Like, now that we have these sensors, what are the other really valuable functionalities that we can provide our customers with through this, of course, taking privacy into account and so on?
Shervin Khodabandeh: This is quite intriguing because Volvo I remember as a child that my uncle used to say, You want a safe car, you get a Volvo. And its always been synonymous with safety, and its really amazing to step back and think about for a company who has put one of its main goals for safe experience, now, with the availability of this amount of data and all of this massive amount of processing, I could imagine there are so many use cases that, to Sams point, are being thought about. So thats really, really encouraging.
What [does] the road map here look like? I mean, is this a constant sort of innovation ideation approach going on to say, What else could we do in these three pillars of personal, safe, and sustainable? What is the process for coming up with these ideas and picking up the good ones and pursuing them or not?
Anders Sjgren: I think that those ideas can either come from the technology side and be really inspired by that, or it could come from the analytics side. And often its when the new technology and the really customer-centric needs where they meet, and also, of course, where we have the agility in the organization to execute on it. Thats really where we have something thats really fruitful. Its typically a mix of different sources of this: innovations and new directions, I would say.
Shervin Khodabandeh: And theres a mechanism to create this interdisciplinary inspiration in the company?
Anders Sjgren: Yeah, I would say so. I would say that there are both formal mechanisms but also informal mechanisms. [As] a car company, Volvo Cars is not super big. Of course there are also smaller ones, but I think its also an advantage that, I mean, pretty much all the different steps, from product strategy, design, the R&D, engineering, and then the later stages
The headquarters is in Gothenburg [in Sweden], within walking distance. The U.S. department is literally 50 meters from where Im sitting right now, while 50 meters in the other direction, there are the crash test facilities and the safety center.
So, what I want to say with that is that its more easy to get connections and to create this kind of how should I put it? informal innovation activities.
Shervin Khodabandeh: This is where relatively smaller size and colocation really, really helps. To have teams that close to each other.
Sam Ransbotham: Its funny to talk about it being small, because its certainly not a small company.
Shervin Khodabandeh: Everythings relative, yeah.
Sam Ransbotham: Everythings relative, I guess. As were chatting, Im thinking about some of the people weve talked to before, and one of the recurring themes that people have mentioned is, oh, this idea [of] starting with a business problem: You dont have AI and find a place to use it. You start with a business problem and then solve it. But this is kind of a nice mix on that that it sounds like theres a lot that starts with a business problem, but then, interestingly, once these processes are in place, and once these technologies are in place, then there becomes a grassroots innovation to say, All right; how can we use that? And thats an interesting perspective that, I think, hasnt come through strongly or maybe Im forgetting something but it seems like that hasnt come through as strongly. This is a nice mix of that, that maybe works in this size organization this colocated organization.
Anders Sjgren: Yeah, and not least in the prototyping and ideation stages. But, then, of course, before it actually goes into the product, it needs to go through a more thorough review and so on.
Shervin Khodabandeh: Theres been quite a few investments and acquisitions of smaller AI startups and firms by Volvo. Tell us a bit about the overall ecosystem of internal and vendor and partner companies that come together to bring to life some of these AI-enabled ideas that youre talking about. Is it internal? Is it external? Is it a mix? How do you think about the ecosystem?
Anders Sjgren: I would definitely say that its a mix. Some of the things we need to do it ourselves to get the full understanding or where we really want to be in the forefront, and [in] some other areas, we definitely want to partner with other companies that are strong in those areas. Traditionally speaking, a car is a super complex product. It has hundreds or thousands of different parts that all need to come together. And, of course, it is a space where there is traditionally a lot of suppliers supplying different parts.
Lately, we are moving more toward bringing software implementation in-house to increase the speed and agility in the development process.
Sam Ransbotham: That seems particularly complicated in auto manufacturing because if I think about how cars got started, they were independent systems. There was a braking system, and a power train system, and an air conditioner, and an infotainment system, and all these were separate. And thats kind of nice because then we have a certain, different standard than we would have for the infotainment system than we would have for the braking system or at least I hope that there would be. But what youre pointing out is, each one of these may be using sensors that come from a different area, and how the whole car has become more complicated independently, but its also become more complicated cohesively, trying to connect all these parts and have them work together.
And that seems, on the one hand, an opportunity for artificial intelligence but, at the same time, a challenge.
Anders Sjgren: Yeah. Its 100% true. Its both a really big opportunity, but that also means thats really one of the core challenges I mean, how to build the cohesive understanding of both the inside and the outside of the car.
We speak about the customer digital twin and vehicle digital twin and so on, and in some sense, those aspects can mean a lot of different things. But, of course, these different systems that you speak about, they are traditionally in different parts of the company, so that also means that there is a lot of cross-functional collaboration that is needed. But we really need to bridge those kind of organizational borders.
I think that thats really one of the key points that is, in order to get successful adoption of the analytics and AI, it really means that different parts of the company need to work together to make it happen. Because otherwise, it will just become a silo. Some people will not really have the benefits of it.
Shervin Khodabandeh: You know, its very true. Its a common theme, and in our work at BCG, we have this rule of thumb the 10:20:70 where we say 10% is the data and algorithms, and 20% is the technology and the digital platform, but 70% of it is the business integration and implementation and bringing different parts of the organization together. This, perhaps, is nowhere more true than at a car company, where you have, as you said, Sam, seemingly disparate systems that are coming together to create a bigger system, but each one of these units has been perfected individually, and now you want the collective perfection as well.
Sam Ransbotham: Anders, we talked a lot about Volvo. How did you personally get interested in artificial intelligence, in data, in technology and analytics? Whats your origin story?
Anders Sjgren: I think Ive always been interested in computers. My father was really an addict, so I kind of grew up with that.
I started out as an engineering student [and got a] masters in computer science. And then I started off as a research engineer in the medical area, at the university hospital in Gothenburg, then found out fairly quickly that in order to really make use of data conclusions and so on come from data so I then went into the area of mathematical statistics.
So I did a Ph.D. in that, then went back to software product development. After some time, I went back to academia for a couple of years, did a postdoc, and then was offered a good opportunity at Volvo Cars. So Ive essentially been here for seven years now. Thats a bit of my history.
Shervin Khodabandeh: Quite inspiring.
Sam Ransbotham: Anders, we also want to ask you a few rapid-fire questions, and the idea is just to answer it as quickly as you can. These are not particularly Volvo questions.
What have you been proudest of that youve done with artificial intelligence?
Anders Sjgren: The problem is that most of the things, I cant speak about.
Shervin Khodabandeh: Thats a great answer too.
Sam Ransbotham: You have to wait and see. OK, well, what worries you about artificial intelligence?
Anders Sjgren: The worries?
Sam Ransbotham: Mm-hmm.
Anders Sjgren: Oh, I think, of course, one is the longer term the kind of singularity things. But I think a bit closer to now, so to speak, we definitely see that this super-fast progression of large language [models] and what they can do, and also the kind of systems that dont just take one problem and give one answer but can really produce a series of steps, in sequence. And that is a super-powerful technology, but a super-powerful technology can be used both for good and for bad.
Thats both something that makes me super excited but also a little bit worried. What will the world look like in 20 years?
Sam Ransbotham: Whats your favorite activity that does not involve technology?
Anders Sjgren: Motorcycling, but obviously using technology. But its more of a yeah.
Sam Ransbotham: Technologys involved. Everythings involved in technology in some way.
Anders Sjgren: Yeah, pretty much.
Sam Ransbotham: Whats the first career you wanted when you were a kid? What did you want to be when you grew up?
Anders Sjgren: Medical doctor.
Sam Ransbotham: Well, that ties with your first career in working in the medical company, then.
Anders Sjgren: Yeah, I think so, but then I found mathematics and those things to be super exciting, so I went into that area.
Sam Ransbotham: Whats your greatest wish for artificial intelligence in the future? What do you hope that we can gain from the advent of these technologies?
Anders Sjgren: If we say the greatest wish, I think it is that we find a way to use it in a way for our common good. We need to find a way to integrate it into society, and I think that that is really my biggest wish for it.
Shervin Khodabandeh: In a relatively short time, we learned so much about various uses of data, AI, and technology, in just what it takes to build a car in new ways, and all the different ways that AI and tech are helping and serving the people who are driving them. Its been really enlightening, Anders. Thank you for joining.
Anders Sjgren: Thanks for having me.
Shervin Khodabandeh: Thanks for listening. On our next episode, Sam and I speak with Shilpa Prasad, entrepreneur in residence at LG Nova. Please join us.
Allison Ryder: Thanks for listening to Me, Myself, and AI. We believe, like you, that the conversation about AI implementation doesnt start and stop with this podcast. Thats why weve created a group on LinkedIn specifically for listeners like you. Its called AI for Leaders, and if you join us, you can chat with show creators and hosts, ask your own questions, share your insights, and gain access to valuable resources about AI implementation from MIT SMR and BCG. You can access it by visiting mitsmr.com/AIforLeaders. Well put that link in the show notes, and we hope to see you there.
Sam Ransbotham (@ransbotham) is a professor in the information systems department at the Carroll School of Management at Boston College, as well as guest editor for MIT Sloan Management Reviews Artificial Intelligence and Business Strategy Big Ideas initiative. Shervin Khodabandeh is a senior partner and managing director at BCG and the coleader of BCG GAMMA (BCGs AI practice) in North America. He can be contacted at shervin@bcg.com.
Me, Myself, and AI is a collaborative podcast from MIT Sloan Management Review and Boston Consulting Group and is hosted by Sam Ransbotham and Shervin Khodabandeh. Our engineer is David Lishansky, and the coordinating producers are Allison Ryder and Sophie Rdinger.
The rest is here:
Fueling Interdisciplinary Innovation With AI: Volvo's Anders Sjgren - MIT Sloan Management Review