Cloud Hosting

Image Credits: Kirillm / Getty Images

Will AI automate human jobs, and if so which jobs and when?

Thats the trio of questions a new research study from MITs Computer Science and Artificial Intelligence Laboratory (CSAIL), out this morning, tries to answer.

Theres been many attempts to extrapolate out and project how the AI technologies of today, like large language models, might impact peoples livelihoods and whole economies in the future.

Goldman Sachs estimates that AI could automate 25% of the entire labor market in the next few years. According to McKinsey, nearly half of all work will be AI-driven by 2055. A survey from the University of Pennsylvania, NYU and Princeton finds that ChatGPT alone could impact around 80% of jobs. And a report from the outplacement firm Challenger, Gray & Christmas suggests that AI is already replacing thousands of workers.

But in their study, the MIT researchers sought to move beyond what they characterize as task-based comparisons and assess how feasible it is that AI will perform certain roles and how likely businesses are to actually replace workers with AI tech.

Contrary to what one (including this reporter) might expect, the MIT researchers found that the majority of jobs previously identified as being at risk of AI displacement arent, in fact, economically beneficial to automate at least at present.

The key takeaway, says Neil Thompson, a research scientist at MIT CSAIL and a co-author on the study, is that the coming AI disruption might happen slower and less dramatically than some commentators are suggesting.

Like much of the recent research, we find significant potential for AI to automate tasks, Thompson told TechCrunch in an email interview. But were able to show that many of these tasks are not yet attractive to automate.

Now, in an important caveat, the study only looked at jobs requiring visual analysis that is, jobs involving tasks like inspecting products for quality at the end of a manufacturing line. The researchers didnt investigate the potential impact of text- and image-generating models, like ChatGPT and Midjourney, on workers and the economy; they leave that to follow-up studies.

In conductingthis study, the researchers surveyed workers to understand what an AI system would have to accomplish, task-wise, to fully replace their jobs. They then modeled the cost of building an AI system capable of doing all this, and also modeled whether businesses specifically non-farm U.S.-based businesses would be willing to pay both the upfront and operating expenses for such a system.

Early in the study, the researchers give the example of a baker.

A baker spends about 6% of their time checking food quality, according to the U.S. Bureau of Labor Statistics a task that could be (and is being) automated by AI. A bakery employing five bakers making $48,000 per year could save $14,000 were it to automate food quality checks. But by the studys estimates, a bare-bones, from-scratch AI system up to the task would cost $165,000 to deploy and $122,840 per year to maintain . . . and thats on the low end.

We find that only 23% of the wages being paid to humans for doing vision tasks would be economically attractive to automate with AI, Thompson said. Humans are still the better economic choice for doing these parts of jobs.

Now, the study does account for self-hosted, self-service AI systems sold through vendors like OpenAI that only need to be fine-tuned to particular tasks not trained from the ground up. But according to the researchers, even with a system costing as little as $1,000, theres lots of jobs albeit low-wage and multitasking-dependent that wouldnt make economic sense for a business to automate.

Even if we consider the impact of computer vision just within vision tasks, we find that the rate of job loss is lower than that already experienced in the economy, the researchers write in the study. Even with rapid decreases in cost of 20% per year, it would still take decades for computer vision tasks to become economically efficient for firms.

The study has a number of limitations, which the researchers to their credit admit. For example, it doesnt consider cases where AI can augment rather than replace human labor (e.g., analyze an athletes golf swing) or create new tasks and jobs (e.g., maintaining an AI system) that didnt exist before. Moreover, it doesnt factor in all the possible cost savings that can come from pre-trained models like GPT-4.

One wonders whether the researchers mightve felt pressure to reach certain conclusions by the studys backer, the MIT-IBM Watson AI Lab. The MIT-IBM Watson AI Lab was created with a $240 million, 10-year gift from IBM, a company with a vested interest in ensuring that AIs perceived as nonthreatening.

But the researchers assert this isnt the case.

We were motivated by the enormous success of deep learning, the leading form of AI, across many tasks and the desire to understand what this would mean for the automation of human jobs, Thompson said. For policymakers, our results should reinforce the importance of preparing for AI job automation...But our results also reveal that this process will take years, or even decades, to unfold and thus that there is time for policy initiatives to be put into place. For AI researchers and developers, this work points to the importance of decreasing the costs of AI deployments and of increasing the scope of how they can be deployed. These will be important for making AI economically attractive for firms to use for automation.

The rest is here:

New MIT CSAIL study suggests that AI wont steal as many jobs as expected - TechCrunch

A cyberattack hit Carnegie Mellon University last summer and the attackers breached personal data, according to a disclosure from the school last week. The Pittsburgh-based university known for its top tech and computer science programs said on Friday that the attack impacted 7,300 students, employees, contractors and other affiliates.

"There is no evidence of fraud or inappropriate use of the information from those files," a statement from CMU said. Still, the attackers likely accessed and copied data that included names, social security numbers and birth dates. With help from law enforcement, CMU disabled any access to that copied data, according to the school.

It started on August 25 when unauthorized users accessed CMU's systems. The university says it began recovery processes and an investigation into the incident that included months later in December, while notifications to impacted parties began to go out last week. Impacted parties will receive credit monitoring services to mitigate further damage.

CMU did not respond to a request for comment and further information about the attack by the time of publication.

Excerpt from:

Carnegie Mellon reveals it was hit by a cyberattack over the summer - Engadget

ATLANTA Armin Mikler has been interested in disaster and emergency response since Hurricane Katrina devasted the Gulf Coast in 2005, killing more than 1,800 people, causing more than $100 billion in damage and exposing serious flaws in the nations ability to respond to disasters.

Mikler, chair of the Department of Computer Sciences at Georgia State University, began working with colleagues at the University of North Texas to develop tools to improve disaster planning and response.

Mikler and his group have developed an algorithm, called the receiving-staging-storing-distributing (RSSD) algorithm, to help public health agencies and others develop a fast and effective response, whether theyre dealing with a hurricane or anthrax attack. In recent tests, they confirmed that it was faster and, in many situations, more effective in helping responders get critical supplies where theyre most needed.

In emergency situations, the population in the affected area needs to be essentially divided up so that medication and other resources can be distributed effectively. This requires the creation of drop points, places in the affected area where supplies are delivered from a central point, or depot, such as the Strategic National Stockpile. The number of vehicles needed to deliver supplies, as well as their carrying capacities, are also major factors. When given the capacity of the vehicles and a time limit, the RSSD algorithm can work out ideal routes to points of delivery.

The question needs to be answered, how do we get from the central point where they drop off, how do we deliver it to all the points where its actually needed? Mikler said.

That depends on how many points we have. And that is a very fluid problem, in the case of such emergencies. For instance, we dont know exactly how many points of dispensing would actually be opening.

Because of the fluidity of these situations, the algorithm he developed needed to be fast to respond effectively to the rapidly changing situations. Usually, problems like this would be solved using an algorithm that provides the best possible and most efficient solution, known as an optimization algorithm. However, optimization algorithms take a long time to find an answer.

This is time that we often do not have when we need to reconfigure our plans, Mikler said.

In short, Mikler and his colleagues found that fast and good enough is better in an emergency than taking too long to find a perfect solution.

To see how this algorithm stacked up to the others in both speed and accuracy, Mikler and Ph.D. student Emma McDaniel conducted benchmarking tests with the RSSD algorithm and others.

Mikler and McDaniel recently published the results of these experiments in the article, Benchmarking a fast, satisficing vehicle routing algorithm for public health emergency planning and response: Good Enough for Jazz. They found that even though the RSSD algorithm doesnt find the optimal solution, it does find consistently good solutions that take a minimum amount of response time. So, not only is the algorithm itself fast, but it also finds some of the fastest routes for resource delivery.

To benchmark the results, Mikler and McDaniel used a database called the CVRPLIB (Capacitated Vehicle Routing Problem Library) as their baseline for best answers to emergency situations. The database contains optimal vehicle route distances for various combinations of depot locations and the number of people who need supplies. Using these datasets, Mikler and McDaniel compared the RSSD algorithm to three others that solve similar problems. In terms of consistency, RSSD came out on top.

The algorithm, first developed in 2014, has been improved over time and is now integrated into response planning software that is used by the Texas Department of State Health Services to assist in both emergency response planning as well as real-time emergency response. Sampson Akwafuo, an assistant professor at California State University, has also used the algorithm to help plan emergency resource delivery in resource-poor areas in some African countries.

Were really able to come up with workable, feasible solutions to problems in a much shorter time, Mikler said.

And time is of the essence in disaster and emergency response.

By Katherine Duplessis

Visit link:

Researcher Develops Computer Algorithm for Disaster Planning and Response - Georgia State University News

A college student has revealed how he was able to stay focused and motivated during an intense period in which he applied for 456 internships.

Oliver Wu is a junior studying computer science at the University of Michigan. He plays an active role in campus life, where he's involved with the university's Asian American community and plays volleyball.

Over the past four months though, Wu has been focused on another task on top of his studies and extracurricular activities: landing an internship.

College students today are increasingly mindful of their future career prospects. A survey of students due to graduate in 2023 conducted by job website Handshake found around half were planning to apply to more jobs, while one third were looking at a more diverse range of roles, and one fifth were starting their search sooner.

Wu has only just started out as a junior, but he's already thinking about the future. He told Newsweek he is seeking a "career in tech," though remains flexible about where that will take him.

"I would love to use my skills to develop solutions in sustainability and environmental protection," he said. "However, I do recognize that I am currently still a college student so there is a lot to learn and my plans may change once I get more experience in the industry."

That desire to seek experience has seen him embark on an exhaustive search to land an internship with a top company. It's a search that began before he even started at college, when he started noticing openings being posted on online job boards over the summer.

"I started applying in July and soon I hit 200 applications," he said. Wu said he "stopped looking at the number of applications" fairly early into the process, but quickly developed a daily routine.

"Usually, I would open up two or three job boards, see what new jobs were posted, and then apply to all of the jobs if the salary, location, roles etc. met what I was looking for," he said. "I also kept track of which companies I had referrals to and checked on a weekly basis if those companies had opened up their applications."

Wu said prior to starting college on his very best days he was completing "15 to 20 applications a day," but that slowed down once he was in class. "On days where I did not apply as much, I would practice my technical interview skills," he added.

He insists he never set out to apply for 456 internships though. "It just kind of happened after applying day in day out," he said. Wu attributes that to the fact that he started applying earlier and continued until late in the year.

During this period, there were times when he felt "burned out" though. "The hardest part was staying positive and working hard, despite having hundreds of rejections," he said.

In those periods, he always made sure to take time off to recharge. He remained motivated though. "I did not want to feel regret that I could have tried harder, so I made up my mind to pursue this with everything I had," he said. He attributes some of that to his religious faith and the belief that whether he succeeded or failed "God has a plan for me."

That plan saw Wu complete an astonishing 56 interviews, as well as 30 technical assessments, 22 second- and third-round assessments and four final rounds off the back of those initial applications.

As stressful as it might have been, those interviews and assessments have proven invaluable. "I feel much less nervous and familiar with the process. Additionally, I know what to expect, and the areas which I need to improve at," he said.

More importantly, at the end of such an intense period of pressure, Wu had something to celebrate. "I ended up accepting an offer at Ford as an enterprise technology intern," he revealed. Wu said the moment he learned he had landed the internship he "felt like a massive weight had been lifted off my shoulders."

"I was in class at the time and I remember stepping out, going into the hallway and jumping up and down while silently screaming in excitement for around 10 minutes," he said. "I ended up landing two more offers, but ultimately accepted Ford."

Eager to share his news and highlight the work that went into it, Wu posted a video to TikTok under the handle oliesandroid revealing how the 456 applications had ultimately been "worth it" in the end. The video has been watched 2.7 million times.

Reflecting on the experience, Wu has one piece of advice for anyone looking to land an internship. "Network," he said. "A big mistake I made was not networking properly and being scared to network and relying on cold applications instead. If I could do it all over again, I would definitely network more. Take a deep breath, and relax, this is a marathon not a sprint."

Newsweek is committed to challenging conventional wisdom and finding connections in the search for common ground.

Newsweek is committed to challenging conventional wisdom and finding connections in the search for common ground.

");jQuery(this).remove()})jQuery('.start-slider').owlCarousel({loop:!1,margin:10,nav:!0,items:1}).on('changed.owl.carousel',function(event){var currentItem=event.item.index;var totalItems=event.item.count;if(currentItem===0){jQuery('.owl-prev').addClass('disabled')}else{jQuery('.owl-prev').removeClass('disabled')}if(currentItem===totalItems-1){jQuery('.owl-next').addClass('disabled')}else{jQuery('.owl-next').removeClass('disabled')}})}})})

Read this article:

College Student Sends 456 Applications, Gets Accepted Into One Internship - Newsweek