11–20 of 91 posts
Must feel pretty bad being a rank and file dev at DeepMind and realizing that your leads had all the talent and funding in the world that they could have used for building language model products and instead they heavily invested in solving board games. Then you realize they get paid >x10 of what you are and they're fine, but the next layoff is likely gonna get you.
Must feel pretty bad being a rank and file dev at DeepMind and realizing that your leads had all the talent and funding in the world that they could have used for building language model products and instead they heavily invested in solving board games. Then you realize they get paid >x10 of what you are and they're fine, but the next layoff is likely gonna get you.
I can't help but wonder when this stuff happens that the devs (or the leads) are worried about being let go, so they open source their stuff to be able to take it elsewhere. Anyone here from DeepMind?
These particular programs were bleeding edge when they were announced, but have been recreated and improved on outside of DeepMind since then. For example, LCZero today is stronger at chess than any of the DeepMind chess programs that were shown to the public. Of course we don't know if they got further behind the scenes. Of course LCZero and everything else relied on the published papers of DeepMind. That is, the pa…
That said many have been retooled for a more specific purpose before public release.
Must feel pretty bad being a rank and file dev at DeepMind and realizing that your leads had all the talent and funding in the world that they could have used for building language model products and instead they heavily invested in solving board games. Then you realize they get paid >x10 of what you are and they're fine, but the next layoff is likely gonna get you.
Humans are also trained in games when young, and even into adulthood, think war games that are held semi-annually between countries.
These are just board games with increased stakes and additional variables.
If you have the research paper, someone in the field could reimplement them in a few days.
Then there is the large compute cost for training them to produce the trained weights.
So, opensourcing these bits of work without the weights isn't as major a thing as you might imagine.
Obligatory note that AlphaZero has long since been surpassed by its independently developed cousin Leela Chess Zero(which is also open source btw), and also Stockfish has more than caught up and remains competitive with MCTS engines.
Obligatory note that AlphaZero has long since been surpassed by its independently developed cousin Leela Chess Zero(which is also open source btw), and also Stockfish has more than caught up and remains competitive with MCTS engines.
Did stockfish use any tech from alphazero & co ? or has it managed to catch up keeping itself "pure" ?
I can't help but wonder when this stuff happens that the devs (or the leads) are worried about being let go, so they open source their stuff to be able to take it elsewhere. Anyone here from DeepMind?
These particular programs were bleeding edge when they were announced, but have been recreated and improved on outside of DeepMind since then. For example, LCZero today is stronger at chess than any of the DeepMind chess programs that were shown to the public. Of course we don't know if they got further behind the scenes. Of course LCZero and everything else relied on the published papers of DeepMind. That is, the pa…
I guess it's because AZ came out of Google?