AI’s progress isn’t the identical as creating human intelligence in machines

But GPT-3 suffers from a number of issues that researchers are working to deal with. It’s usually inconsistent—you will get contradictory solutions to the identical query. Second, GPT-3 is liable to “hallucinations”: when requested who the president of the United States was in 1492, it’s going to fortunately conjure up a solution. Third, GPT-3 is an costly mannequin to coach and costly to run. Fourth, GPT-3 is opaque—it’s obscure why it drew a specific conclusion. Finally, since GPT-3 parrots the contents of its coaching knowledge, which is drawn from the net, it usually spews out poisonous content material, together with sexism, racism, xenophobia, and extra. In essence, GPT-3 can’t be trusted.

Despite these challenges, researchers are investigating multi-modal variations of GPT-3 (corresponding to DALL-E2), which create lifelike photographs from natural-language requests. AI builders are additionally contemplating find out how to use these insights in robots that work together with the bodily world. And AI is more and more being utilized to biology, chemistry, and different scientific disciplines to glean insights from the large knowledge and complexities in these fields. 

The bulk of the fast progress at the moment is on this data-centric AI, and the work of this 12 months’s 35 Innovators Under 35 winners isn’t any exception. While data-centric AI is highly effective, it has key limitations: the programs are nonetheless designed and framed by people. Just a few years in the past, I wrote an article for MIT Technology Review referred to as “How to know if synthetic intelligence is about to destroy civilization.” I argued that efficiently formulating issues stays a distinctly human functionality. Pablo Picasso famously stated, “Computers are ineffective. They solely provide you with solutions.” 

We proceed to anticipate the distant day when AI programs can formulate good questions—and shed extra gentle on the elemental scientific problem of understanding and establishing human-level intelligence. 

Oren Etzioni is CEO of the Allen Institute for AI and a choose for this 12 months’s 35 Innovators competitors.