Researchers emphasized the rapid pace of the AI ​​industry


As outside observers, Ai researchers are in an unenviable position. They're the big tech companies they're looking for. They are shooting an eye at home. They are in the hottest industry right now.

But all this comes with intense pressure.

More than a dozen dragon warriors who spoke to researchers about the case, some of whom spoke on condition of anonymity for fear of retribution, said Breakneck Pace had affected their mental health. Competition They say it creates an atmosphere of isolation among AI Labs.

“Everything has changed overnight,” said one researcher, “combining our work with positive and negative outcomes.”

This past December, Openai 12 livestreams 12 hosted At that time it announced more than 0 new devices, models and services. Google Response versus toolsand models Services Its own press releases, background and exit between the two tech giants is unique for its speed.

The grind and the hustle

Silicon Valley is no stranger to hustle culture. But with the rise of AI, public support for overwork has become a hindrance.

At Openai No Unique Time well spent with researchers to work six days a week. CEO Sam Altman told the company's teams to push Turn achievements into Public product On a rough schedule. Bob McGrew, Chief Research Officer of Openai, Report It was cited as one of the reasons for his departure last September.

None of the comforts found in competing labs. Google Bladmind Team Google's AI models Google's flagship Gemini Developmind Gemini works from a few hours to 100 hours at a time Bug fix for the system. Engineers at XAI, Elon Musk's AI company; Regularly Pillar About nights of work that bleed to worms in the morning.

Why push relentlessly? AI research today can have a huge impact on a company's bottom line. Google Parent Script which is growing Google's GoGini Champokot, Google's GoGini Champokot, is to reveal the controversial renderings of historical figures.

“One of the biggest pressures is competitiveness,” says KAI Arulkumaran, research lead at Ai Services provider Araya.

All above the leadership board

Some of these contests are highly publicized.

On a monthly and sometimes weekly basis, AI companies are trading one-to-one to replace one-to-one AI models such as the chatbot arena such as the Compbot arena. Logan Kilpatrick, Product Lead for Google Gemini Goodi Developer Tools I admit it In a manuscript on X, the Chatbot Arena has a disparity effect on the velocity of AI development. “

Not all researchers believe this is a good thing. Such is the speed of the industry that they can find their work before they even get sent out.

“It makes me question the value of their work,” said Zihan Wang, a robotics engineer who works at a stealth AI startup.

Other researchers serve to focus attention on trade fairs.

“On top of the stress (causes of stress) is the transition for AI researchers to work on (AI models) in-house programs in industries and deliver solutions for products,” Arikumaran said. “Industry has built up an expectation that AI researchers can pursue academic research in industry, but this is no longer the case.

Another researcher is their worries and anxiety.

“Now, before going back to the scientific community, a commercial resource,

Run to Grauntlet

Some researchers trace the seeds of their concern to AI grad programs.

Gowthami Apalli, a PhD student studying AI at the University of Maryland, said that research is being published so quickly that, in some cases, many have seen AI companies prefer candidates with highly relevant experience.

A PHD is generally a lonely and stressful experience, and a PHD can be especially stressful when you're publishing 4 papers a year while the students in your lab are publishing only 1 or 2 papers a year, with the rapid growth of the Field and the 'spirit that's either broadcast or destroyed'. “

Some say that after the first two years of the literal program, she stopped taking vacations because she felt guilty before studying.

“During my PhD I was constantly ill and lost it at the end of my first year,” he said.

The way forward

So what changes, if. What changes can AI make to the work environment? Slowing down the pace of development does not come with many risks.

Some small but impactful adjustments are emphasized, such as regularly expressing one's personal challenges.

“One of the biggest problems is that no one is openly discussing their struggles, everyone puts on a brave face. I believe (people) would feel better if they could see that others are struggling.”

Bhaskar Bhatt, AI consultant at professional services firm EY, said the industry should work to build “robust support networks” to combat feelings of isolation.

“Fostering a culture that values ​​work-life balance is essential, where individuals can truly connect from their work,” Bhattat said.

Raj Dabre, a postdoCtoral student at Princeton and an AI researcher at the National Institute of Information and Communications Technology in Japan, said in a press release that researchers should be reminded in gentler ways.

“We need to say that AI is work, starting with people.”



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *