3 ms·
I'd expect it's more that Microsoft probably shares my perspective - that the current direct and path of technology will never produce AGI, so the distinction i
by somenameforme 3y ago
I'd expect it's more that Microsoft probably shares my perspective - that the current direct and path of technology will never produce AGI, so the distinction is largely irrelevant, and makes the investment sound more societally friendly.
- rational_indian 3y agoIt is only a matter of time before we have true AGI. It's a hardware problem. We don't need to wait for someone to invent a clever algorithm for this. OpenAI knows this. See... - http://www.incompleteideas.net/IncIdeas/BitterLesson.html http://www.incompleteideas.net/IncIdeas/BitterLesson.html - https://en.wikipedia.org/wiki/Huang%27s_law https://en.wikipedia.org/wiki/Huang%27s_law
- pc86 3y agoIf you're going the argument from authority route you should probably cite sources that don't say "The hypothesis is subject to questions about its validity." in the very first paragraph.
- rational_indian 3y agoEvery hypothesis is subject to questions about its validity. Including Moore's law but people still use it.
- pc86 3y agoSure, so is gravity. The Wikipedia page on gravity doesn't include such a disclaimer.
- notahacker 3y agoAnd frankly even if they did produce something the board deemed to be AGI, I don't imagine the integration of LLMs and generative art into Microsoft products beforehand wouldn't have been worth the investment
- wslh 3y agoYes, I interpret as a maximum that this clause about AGI is just a bargaining chip or a cosmetic point: I renounce all my rights if we have a datacenter in Mars. Clearly it shows the vision about AGI. Another interpretation is that if AGI is achieved it will be open available beyond OpenAI and accessed through other sources.