Cloud Hosting

Readers express their hopes, and fears, about recent developments in artificial intelligence chatbots

Evgeny Morozovs piece is correct insofar as it states that AI is a long way from the general sentient intelligence of human beings (The problem with artificial intelligence? Its neither artificial nor intelligent, 30 March). But that rather misses the point of the thinking behind the open letter of which I and many others are signatories. ChatGPT is only the second AI chatbot to pass the Turing test, which was proposed by the mathematician Alan Turing in 1950 to test the ability of an AI model to convincingly mimic a conversation well enough to be judged human by the other participant. To that extent, current chatbots represent a significant milestone.

The issue, as Evgeny points out, is that a chatbots abilities are based on a probabilistic prediction model and vast sets of training data fed to the model by humans. To that extent, the output of the model can be guided by its human creators to meet whatever ends they desire, with the danger being that its omnipresence (via search engines) and its human-like abilities have the power to create a convincing reality and trust where none does and should exist. As with other significant technologies that have had an impact on human civilisation, their development and deployment often proceeds at a rate far faster than our ability to understand all their effects leading to sometimes undesirable and unintended consequences.

We need to explore these consequences before diving into them with our eyes shut. The problem with AI is not that it is neither artificial nor intelligent, but that we may in any case blindly trust it.Alan LewisDirector, SigmaTech Analysis

The argument that AI will never achieve true intelligence due to its inability to possess a genuine sense of history, injury or nostalgia and confinement to singular formal logic overlooks the ever-evolving capabilities of AI. Integrating a large language model in a robot would be trivial and would simulate human experiences. What would separate us then? I recommend Evgeny Morozov watch Ridley Scotts Blade Runner for a reminder that the line between man and machine may become increasingly indistinct. Daragh ThomasMexico City, Mexico

Artificial intelligence sceptics follow a pattern. First, they argue that something can never be done, because it is impossibly hard and quintessentially human. Then, once it has been done, they argue that it isnt very impressive or useful after all, and not really what being human is about. Then, once it becomes ubiquitous and the usefulness is evident, they argue that something else can never be done. As with chess, so with translation. As with translation, so with chatbots. I await with interest the next impossible development.Edward HibbertChipping, Lancashire

AIs main failings are in the differences with humans. AI does not have morals, ethics or conscience. Moreover, it does not have instinct, much less common sense. Its dangers in being subject to misuse are all too easy to see.Michael ClarkSan Francisco, US

Thank you, Evgeny Morozov, for your insightful analysis of why we should stop using the term artificial intelligence. I say we go with appropriating informatics instead.Annick DriessenUtrecht, the Netherlands

{{topLeft}}

{{bottomLeft}}

{{topRight}}

{{bottomRight}}

{{.}}

View original post here:
The danger of blindly embracing the rise of AI - The Guardian

Last week, a whos who of technologists called for artificial intelligence (AI) labs to stop training the most powerful AI systems for at least six months, citing "profound risks to society and humanity."

In an open letter that now has more than 3,100 signatories, including Apple co-founder Steve Wozniak, tech leaders called out San Francisco-based OpenAI Labs recently announcedGPT-4 algorithm in particular, saying the company should halt further development until oversight standards are in place. That goal has the backing of technologists, CEOs, CFOs, doctoral students, psychologists, medical doctors, software developers and engineers, professors, and public school teachers from all over the globe.

On Friday, Italy became the first Western nation to ban further development of ChatGPT over privacy concerns; the natural language processing app experienced a data breach last month involving user conversations and payment information. ChatGPT is the popular GPT-based chatbot created by OpenAI and backed by billions of dollars from Microsoft.

The Italian data protection authority said it is also investigating whether OpenAI's chatbot already violated the European Union'sGeneral Data Protection Regulationrules created toprotect personal data inside and outside the EU.OpenAI has complied with the new law, according to a report by the BBC.

The expectation among many in the technology community is that GPT, which stands for Generative Pre-trained Transformer, will advance to become GPT-5 and that version will be an artificial general intelligence, or AGI. AGI represents AI that can think for itself, and at that point, the algorithm would continue to grow exponentially smarter over time.

Around 2016, a trend emerged in AI training models that were two-to-three orders of magnitude larger than previous systems, according to Epoch,a research group trying to forecast the development of transformative AI. That trend has continued.

There are currently no AI systems larger than GPT-4 in terms of training compute, according to Jaime Sevilla, director of Epoch. But that will change.

Large-scale Machine Learning models for AI have more than doubled in capacity ever year.

Anthony Aguirre, a professor of physics at UC Santa Cruz and executive vice president of the Future of Life, the non-profit organization that published the open letter to developers, said theres no reason to believe GPT-4 wont continue to more than double in computational capabilities every year.

The largest-scale computations are increasing size by about 2.5 times per year. GPT-4s parameters were not disclosed by OpenAI, but there is no reason to think this trend has stopped or even slowed, Acquirre said. Only the labs themselves know what computations they are running, but the trend is unmistakable.

In his biweekly blog on March 23, Microsoft co-founder Bill Gates heralded AGI which is capable of learning any task or subject as the great dream of the computing industry.

AGI doesnt exist yet there is a robust debate going on in the computing industry about how to create it, and whether it can even be created at all, Gates wrote. Now, with the arrival of machine learning and large amounts of computing power, sophisticated AIs are a reality, and they will get better very fast.

Muddu Sudhakar, CEO ofAisera, agenerative AI company for enterprises, saidthere are but a handful of companies focused on achieving AGI as OpenAI and DeepMind (backed by Google), though they have "huge amounts of financial and technical resources."

Even so, they have a long way to go to get to AGI, he said.

"There are so many tasks AI systems cannot do that humans can do naturally, like common-sense reasoning, knowing what a fact is and understanding abstract concepts (such as justice, politics, and philosophy)," Sudhakar said in an email to Computerworld. "There will need to be many breakthroughs and innovations for AGI. But if this is achieved, it seems like this system would mostly replace humans.

"This would certainly be disruptive and there would need to be lots of guardrails to prevent the AGI from taking full control," Sudhakar said. "But for now, this is likely in the distant future. Its more in the realm of science fiction."

Not everyone agrees.

AI technology and chatbot assistants have and will continue to make inroads in nearly every industry. The technology can create efficiencies and take over mundane tasks, freeing up knowledge workers and others to focus on more important work.

For example, large language models (LLMs) the algorithms powering chatbots can sift through millions of alerts, online chats, and emails, as well as finding phishing web pages and potentially malicious executables. LLM-powered chatbots can write essays, marketing campaigns and suggest computer code, all from just simple user prompts (suggestions).

Chatbots powered by LLMs are natural language processors that basically predict the next words after being prompted by a users question. So, if a user were to ask a chatbot to create a poem about a person sitting on a beach in Nantucket, the AI would simply chain together words, sentences and paragraphs that are the best responses based on previous training by programmers.

But LLMs also have made high-profile mistakes, and can produce hallucinations where the next-word generation engines go off the rails and produce bizarre responses.

If AI based on LLMs with billions of adjustable parameters can go off the rails, how much greater would the risk be when AI no longer needs humans to teach it, and it can think for itself? The answer is much greater, according to Avivah Litan, a vice president and distinguished analyst at Gartner Research.

Litan believes AI development labs are moving forward at breakneck speed without any oversight, which could result in AGI becoming uncontrollable.

AI laboratories, she argued, have raced ahead without putting the proper tools in place for users to monitor whats going on. I think its going much faster than anyone ever expected, she said.

The current concern is that AI technology for use by corporations is being released without the tools users need to determine whether the technology is generating accurate or inaccurate information.

Right now, were talking about all the good guys who have all this innovative capability, but the bad guys have it, too, Litan said. So, we have to have these water marking systems and know whats real and whats synthetic. And we cant rely on detection, we have to have authentication of content. Otherwise, misinformation is going to spread like wildfire.

For example, Microsoft this week launched Security Copilot, which is based on OpenAIs GPT-4 large language model. The tool is an AI chatbot for cybersecurity experts to help them quickly detect and respond to threats and better understand the overall threat landscape.

The problem is, you as a user have to go in and identify any mistakes it makes, Litan said. Thats unacceptable. They should have some kind of scoring system that says this output is likely to be 95% true, and so it has a 5% chance of error. And this one has a 10% chance of error. Theyre not giving you any insight into the performance to see if its something you can trust or not.

A bigger concern in the not-so-distant future is that GPT-4 creator OpenAI will release an AGI-capable version. At that point, it may be too late to rein in the technology.

One possible solution, Litan suggested, is by releasing two models for every generative AI tool one for generating answers, the other for checking the first for accuracy.

That could do a really good job at ensuring if a model is putting out something you can trust, she said. You cant expect a human being to go through all this content and decide whats true or not, but if you give them other models that are checking, that would allow users to monitor the performance.

In 2022, Time reported that OpenAI had outsourced services to low-wage workers in Kenya to determine whether its GPT LLM was producing safe information. The workers hired by Sama, a San Francisco-based firm, were reportedly paid $2 per hour and required to sift through GPT app responses that were prone to blurting out violent, sexist and even racist remarks.

And this is how youre protecting us? Paying people $2 an hour and who are getting sick. Its wholly inefficient and its wholly immoral, Litan said.

AI developers need to work with policy makers, and these should at a minimum include new and capable regulatory authorities, Litan continued. I dont know if well ever get there, but the regulators cant keep up with this, and that was predicted years ago. We need to come up with a new type of authority.

Shubham Mishra, co-founder & global CEO for AI start-upPixis, believes while progress in his field cannot, and must not, stop, the call for a pause in AI development is warranted. Generative AI, he said, does have the power to confuse masses by pumping out propaganda or "difficult to distinguish" information into the public domain.

What we can do is plan for this progress. This can be possible only if all of us mutually agree to pause this race and concentrate the same energy and efforts on building guidelines and protocols for the safe development of larger AI models, Mishra said in an email to Computerworld.

In this particular case, the call is not for a general ban on AI development but a temporary pause on building larger, unpredictable models that compete with human intelligence, he continued. The mind-boggling rates at which new powerful AI innovations and models are being developed definitely calls for the tech leaders and others to come together to build safety measures and protocols.

Read more here:
Should we fear the rise of artificial general intelligence? - Computerworld

This article has been reviewed according to ScienceX's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:

fact-checked

trusted source

proofread

Credit: Pixabay/CC0 Public Domain

The development of artificial intelligence (AI) is out of control, in the opinion of approximately 3,000 signatories of an open letter published by business leaders and scientists.

The signatories call for a temporary halt to training especially high-performance AI systems. Prof. Urs Gasser, expert on the governance of digital technologies, examines the important questions from which the letter deflects attention, talks about why an "AI technical inspection agency" would make good sense and looks at how far the EU has come compared to the U.S. in terms of regulation.

Artificial intelligence systems capable of competing with human intelligence may entail grave risks for society and humanity, say the authors of the open letter. Therefore, they continue, for at least six months no further development should be conducted on technologies which are more powerful than the recently introduced GPT-4, successor to the language model ChatGPT.

The authors call for the introduction of safety rules in collaboration with independent experts. If AI laboratories fail to implement a development pause voluntarily, governments should legally mandate the pause, says the signatories.

Unfortunately the open letter absorbs a lot of attention which would be better devoted to other questions in the AI debate. It is correct to say that today probably nobody knows how to train extremely powerful AI systems in such a way that they will always be reliable, helpful, honest and harmless.

Nonetheless, a pause in AI training will not help achieve this, primarily because it would be impossible to assert such a moratorium on a global level, and because it would not be possible to implement the regulations called for within period of only six months. I'm convinced that what's necessary is a stepwise further development of technologies in parallel to the application and adaptation of control mechanisms.

First of all, the open letter once again summons up the specter of what is referred to as an artificial general intelligence. That deflects attention from a balanced discussion of the risks and opportunities represented by the kind of technologies currently entering the market. Second, the paper refers to future successor models of GPT-4.

This draws attention away from the fact that GPT-4's predecessor, ChatGPT, already presents us with essential challenges that we urgently need to addressfor example misinformation and prejudices which the machines replicate and scale. And third, the spectacular demands made in the letter distract us from the fact that we already have instruments now which we could use to regulate the development and use of AI.

Recent years have seen the intensive development of ethical principles which should guide the development and application of AI. These have been supplemented in important areas by technical standards and best practices. Specifically, the OECD Principles on Artificial Intelligence link ethical principles with more than 400 concrete tools.

And the US National Institute of Standards and Technology (NIST) has issued a 70-page guideline on how distortions in AI systems can be detected and handled. In the area of security in major AI models, we're seeing new methods like constitutional AI, in which an AI system "learns" principles of good conduct from humans and can then use the results to monitor another AI application. Substantial progress has been made in terms of security, transparency and data protection and there are even specialized inspection companies.

Now the essential question is whether or not to use such instruments, and if so how. Returning to the example of ChatGPT: Will the chat logs of the users be included in the model for iterative training? Are plug-ins allowed which can record user interactions, contacts and other personal data? The interim ban and the initiation of an investigation of the developers of ChatGPT by the Italian data protection authorities are signs that very much is still unclear here.

The history of technology has taught us that it is difficult to predict the "good" or "bad" use of technologies, even that technologies often entail both aspects and negative impacts can often be unintentional. Instead of fixating on a certain point in a forecast, we have to do two things: First, we have to ask ourselves which applications we as a society do not want, even if they were possible. We need clear red lines and prohibitions.

Here I'm thinking of autonomous weapons systems as an example. Second, we need comprehensive risk management, spanning the range from development all the way to use. The demands placed here increase as the magnitude of the potential risks to people and the environment posed by a given application grow. European legislature is correct in taking this approach.

This kind of independent inspection is a very important instrument, especially when it comes to applications that can have a considerable impact on human beings. And by the way, this is not a new idea: we already see inspection procedures and instances like these at work in the wide variety of aspects of life, ranging from automobile inspections to general technical equipment inspections and financial auditing.

However, the challenge is disproportionally greater with certain AI methods and applications, because certain systems develop themselves as they are used, i.e. they are dynamic in nature. And it's also important to see that experts alone won't be able to make a good assessment of all societal impacts. We also need innovative mechanisms which for example include disadvantaged people and underrepresented groups in the discussion on the consequences of AI. This is no easy job, one I wish was attracting more attention.

We do indeed need clear legal rules for artificial intelligence. At the EU level, an act on AI is currently being finalized which is intended to ensure that AI technologies are safe and comply with fundamental rights. The draft bill provides for the classification of AI technologies according to the threat they pose to these principles, with the possible consequence of prohibition or transparency obligations.

For example, plans include prohibiting evaluation of private individuals in terms of their social behavior, as we are currently seeing in China. In the U.S. the political process in this field is blocked in Congress. It would be helpful if the prominent figures who wrote the letter would put pressure on US federal legislators to take action instead of calling for a temporary discontinuation of technological development.

The rest is here:
A freeze in training artificial intelligence won't help, says professor - Tech Xplore