More AI researchers claim that modern programs are reaching sentience

share to other networks share to twitter share to facebook

Earlier this week, the artificial intelligence community was thrown into an uproar over claims of AI consciousness. AI researchers from around the globe all came out to dismiss all claims towards alleged sentience.

The topic was kicked off with a discussion from OpenAI co-founder Ilya Sutskever. In a single sentence tweet, the AI developer explained that “today’s large neural networks are slightly conscious”.

AI researchers back Sutskever

Advertisement

Sutskever’s claims brought out the ire of almost every artificial intelligence researcher on social media. However, after receiving the fury of the Internet, Sutskever was eventually backed by others in his field.

Via Independent, AI expert Tamay Besiroglu backed Sutskever’s claims that AI could be in the early stages of consciousness. The AI researcher explained that the OpenAI founder’s claims shouldn't be so easily dismissed.

“Seeing so many prominent machine learning folks ridiculing this idea is disappointing,” the expert tweeted. “It makes me less hopeful in the field’s ability to seriously take on some of the profound, weird and important questions that they’ll undoubtedly be faced with over the next few decades.”

Besiroglu’s statement comes after a recent study ranking consciousness in artificial intelligence programs. The study placed programs on a graph that stretched from Not Conscious to Maybe Slightly Conscious.

The Maybe Slightly Conscious category was miniscule. However, OpenAI text generator GPT-3 and Google DeepMind’s AlphaGo Zero could be deemed slightly conscious. But just what does that mean for AI?

PSX 20220218 140411
click to enlarge

Read More: Clearview AI won't stop until every humans face is stored in its database

Advertisement

A new form of consciousness

The sense of consciousness is one that's hard to define in living, flesh-and-blood beings. In artificial intelligence, that sense will be even harder to define, especially as AI programs get increasingly more complex.

OpenAI CEO Sam Altman explained that a consciousness for an artificial intelligence program won't be the same as a living one. While an AI may reach sentience, a self-aware form, it will still have differences.

“I think GPT-3 or -4 will very, very likely not be conscious in any way we use that word,” the CEO said.  “If they are, it’s a very alien form of consciousness.”