Monday, April 11, 2022

Coming Soon: General Artificial Intelligence

The closer you get to experts who understand the nuts and bolts and history of AI, the more you find them saying that what we have is not nearly General Artificial Intelligence (GAI), and that GAI seems far away. I think we already have the roots in place with Neural Networks (NN), Deep Learning (DL), Machine Learning (ML), and primitive domain limited Artificial Intelligence (AI). Things like computer vision, voice recognition, and language translation are already in production. These are tough problems, but in some ways, machines are already better than humans are. I expect GAI to be an emergent property as systems mature, join, and augment one another.

I was around during the 70s AI winter, and was involved in the 80s AI winter as one of the naysayers. I built a demonstration system with a Sperry voice recognition card in 1984. I could demonstrate it in a quiet room, but as a practical matter, it was not production ready at all. Around 1988 we built demonstration expert systems using very expensive tools provided by vendors. At the end of the day, these were also far too limited to use in production.

Until about 2012, I would have said we don't have AI and are not that close. This was more than a decade after a computer had beaten a human grandmaster at chess.

During the last decade or so, pieces have begun to fall into place that make me think machines will soon contribute more to knowledge and understanding than human beings. For about five years or so, I have been saying that improvements in AI will accelerate until we reach something of a singularity. Machines will become entirely more intelligent than human beings are. We will have a positive feedback loop as better AI systems design even better AI systems.

In limited domains like chess, machines beat humans every time. Arguably, they already have a more sophisticated understanding than humans do.

Despite the AI winters, I always thought that machines' becoming better than humans at chess was inevitable. For a long time, some people vigorously disputed this. Now, it’s a given.

I disagreed with experts in the past. I said that what they had was not AI and not likely to become AI. To some extent, I disagree with the experts again. I think that AI is finally real. GAI is inevitable, and now quite close.

No comments:

AI Alignment and Security Now!

As far as I know, people charged with 'Alignment' at AI companies are not convinced things are safe enough and some have quit becaus...