(C) Daily Kos
This story was originally published by Daily Kos and is unaltered.
 . . . . . . . . . .



Man Beats Machine at Go: There is No Such Thing as Artificial Intelligence [1]

['This Content Is Not Subject To Review Daily Kos Staff Prior To Publication.', 'Backgroundurl Avatar_Large', 'Nickname', 'Joined', 'Created_At', 'Story Count', 'N_Stories', 'Comment Count', 'N_Comments', 'Popular Tags']

Date: 2023-02-22  

A man beat a machine at Go, something that hasn't happened in a while:

A human player has comprehensively defeated a top-ranked AI system at the board game Go, in a surprise reversal of the 2016 computer victory that was seen as a milestone in the rise of artificial intelligence. Kellin Pelrine, an American player who is one level below the top amateur ranking, beat the machine by taking advantage of a previously unknown flaw that had been identified by another computer. But the head-to-head confrontation in which he won 14 of 15 games was undertaken without direct computer support. The triumph, which has not previously been reported, highlighted a weakness in the best Go computer programs that is shared by most of today’s widely used AI systems, including the ChatGPT chatbot created by San Francisco-based OpenAI. Man beats machine at Go in human victory over AI | Ars Technica

What is interesting about this is that if they had been playing a person, the tactic would not have worked, at least not against anyone competent:

The tactics used by Pelrine involved slowly stringing together a large “loop” of stones to encircle one of his opponent’s own groups, while distracting the AI with moves in other corners of the board. The Go-playing bot did not notice its vulnerability, even when the encirclement was nearly complete, Pelrine said. “As a human it would be quite easy to spot,” he added. The discovery of a weakness in some of the most advanced Go-playing machines points to a fundamental flaw in the deep-learning systems that underpin today’s most advanced AI, said Stuart Russell, a computer science professor at the University of California, Berkeley. The systems can “understand” only specific situations they have been exposed to in the past and are unable to generalize in a way that humans find easy, he added.

The last line is terribly important. These systems are not intelligent in any meaningful way. They can be taught to do specialized tasks, and there can be value in those tasks. But they cannot learn in the way that humans can -- they cannot take past experience and generalize from it on the fly. They must be trained in very specific, data intensive ways. And if you confront them with material or situations they have not been trained to handle, they fall apart.

This is not just a matter of semantics. Words have meanings, and by labeling these machines intelligent, their creators can convince the general public that a word correlation machine or pixel re-organizer should be allowed to take people's work without compensation. Because, after all, isn't that how people learn, and isn't the machine just learning?

As we can see, even avoiding the tricky question of what a human brain is doing when we learn, a machine isn't learning in any meaningful sense. It is, at best, copying without the ability to understand what it is copying, to understand how to extrapolate and generalize. It hasn't learned how to play Go. It has merely learned how to recognize many, many patterns that occur during Go games. Those are not, I hope you can see, the same things.

Machines are not now intelligent. No matter how much data we through at them, what we have created are not systems that learn but systems that copy. If you happen to encounter a similar pattern, then you do okay. If not, then you lose to an amateur, or tell a reporter that he should get divorced.

And given that reality, when questions of ownership or misappropriation or basic fairness come up, don't you think we should lean a little on the side of the actual intelligence beings?

Want more oddities like this? You can follow my RSS Feed or newsletter.

[END]
---
[1] Url: https://www.dailykos.com/stories/2023/2/22/2154067/-Man-Beats-Machine-at-Go-There-is-No-Such-Thing-as-Artificial-Intelligence

Published and (C) by Daily Kos
Content appears here under this condition or license: Site content may be used for any purpose without permission unless otherwise specified.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/dailykos/