Comment by avs733 | Hacker News Reader

Comment by avs733 | Hacker News Reader

Parent Full thread

Another way to look at your metaphor is that research on exercise physiology has shown the enormous importance of rest and proper nutrition during training. Prior to cars, getting from place to place and access to optimal nutrition were both mediated by transport over long distances.

AI is the future of Go because it enables those new perspectives and new processes by which human players can learn. AI is smart-dumb, looking for patterns beyond the human capacity, but limited by the data on existing human players that has been provided to them.

Strava hasn't made running races pointless. I can compare my runs to others but that is a very different metric then beating them in a head to head race.

> AI is smart-dumb, looking for patterns beyond the human capacity, but limited by the data on existing human players that has been provided to them.

I think you misunderstand how AI in games like go now works. Most of the advancements recently have been from the AI playing itself, oftentimes without any database of human moves at all.

> limited by the data on existing human players that has been provided to them.

Except it isn't. In various games, new strategies have been found by playing AI vs AI. It's also possible to create AI players by self-play with no knowledge of human matches.

Technically AlphaGo and especially AlphaGo Zero create their own training data. So they’re not limited to data from human players.

Do these AIs actually have much of a strategy? Is the strategy mostly correct evaluations of positions and optimized search?

The search space of Go is way too large for dumb traverse of the tree, even with high end optimizations.

What makes recent breakthroughs in AI agents playing adversarial games possible is the fact that deep neural networks are able to develop patterns that yield short- and long-term strategic planning. And the ability to self train without human intervention to reach unprecedented training levels.

Reply on news.ycombinator.com