After many years of academia leading the way in the development of artificial intelligence, the tides have shifted and industry has taken over, according to the 2023 AI Index, a report created by Stanford University with help from companies including Google, Anthropic and Hugging Face. “In 2022, there were 32 significant industry-produced machine learning models compared to just three produced by academia,” the report says. The shift in influence is attributed mainly to the large resource demands — in staff, computing power and training data — required to create state of the art AI systems.

To illustrate that point, The Verge compares the 2019 creation of OpenAI’s 1.5 billion parameter GPT-2, an early large language model that cost about $50,000 to train with the 2022 debut of Google’s PaLM, which “cost an estimated $8 million to train and contains 540 billion parameters, making it 360 times larger than GPT-2 and 160 times more expensive.”

AI tools that have been used by millions of people in the U.S. this past year have included OpenAI’s ChatGPT chatbot, its image generator DALL-E and competing platform Midjourney.

Despite that popularity, the U.S. was among the countries in which citizens felt the least positive about AI, with only 35 percent of respondents agreeing with the statement that “products and services using AI have more benefits than drawbacks,” says the 2023 AI Index Report, prepared by Stanford’s Institute for Human-Centered Artificial Intelligence.

That contrasts to 78 percent in China, 76 percent in Saudi Arabia and 71 percent in India. That seems to indicate American agreement with the scientists and stakeholders calling for a “pause” on giant AI experiments due to concerns that the shift in balance of power toward commercial players may lead to “dangerous outcomes” as business incentives prompt “companies rush out products and sideline safety concerns in an effort to outmaneuver rivals,” the report explains.

As corporations take AI mainstream “the number of incidents of ethical misuse has also increased,” The Verge writes, citing the AIAAIC (AI, Algorithmic, and Automation Incidents and Controversies) Repository notes a 26-fold increase of problems between 2012 and 2021.

“Such incidents include fatalities involving Tesla’s self-driving software; the use of audio deepfakes in corporate scams; the creation of nonconsensual deepfake nudes; and numerous cases of mistaken arrests caused by faulty facial recognition software, which is often plagued by racial biases,” according to The Verge.

TechCrunch — which calls the 386-page report “quite readable and nontechnical” — cites among the noteworthy findings that global private investment in AI declined in 2022 to $91.9 billion, down 26.7 percent from the previous year, “but that’s after an astronomic increase over the last decade.”

“AI will continue to improve and, as such, become a greater part of all our lives,” AI Index steering committee co-directors Jack Clark and Ray Perrault write, adding that “given the increased presence of this technology and its potential for massive disruption, we should all begin thinking more critically about how exactly we want AI to be developed and deployed.”