We remain committed to our partnership with OpenAI and have confidence in our product roadmap, our ability to continue to innovate with everything we announced at Microsoft Ignite, and in continuing to support our customers and partners. We look forward to getting to know Emmett…
— Satya Nadella (@satyanadella) November 20, 2023
We remain committed to our partnership with OpenAI and have confidence in our product roadmap, our ability to continue to innovate with everything we announced at Microsoft Ignite, and in continuing to support our customers and partners. We look forward to getting to know Emmett Shear and OAI's new leadership team and working with them. And we're extremely excited to share the news that Sam Altman and Greg Brockman, together with colleagues, will be joining Microsoft to lead a new advanced AI research team. We look forward to moving quickly to provide them with the resources needed for their success.
Confirmed to be a separate subsidiary entity https://x.com/satyanadella/status/1726516824597258569
I'm super excited to have you join as CEO of this new group, Sam, setting a new pace for innovation. We've learned a lot over the years about how to give founders and innovators space to build independent identities and cultures within Microsoft, including GitHub, Mojang Studios, and LinkedIn, and I'm looking forward to having you do the same.
Elon peepeeriding in replies https://x.com/elonmusk/status/1726540022571299281
Wild times
Satya ftw
Reactions
https://news.ycombinator.com/item?id=38344196
https://old.reddit.com/r/singularity/comments/17zjg2k/sam_antman_and_greg_brockman_join_microsoft
https://old.reddit.com/r/OpenAI/comments/17zjkj4/sam_altman_and_greg_brockman_together_with
https://old.reddit.com/r/technology/comments/17zjtha/sam_altman_greg_brockman_and_colleagues_joins
https://old.reddit.com/r/technology/comments/17zjoe2/microsoft_hires_former_openai_ceo_sam_altman
Jump in the discussion.
No email address required.
that's not what I'm talking about.
you can check their publications in NIPS, ICML, ICLR, JMLR as well as the open source code.
i don't think so. openAI increased the model size and hired a lot of people to improve the model afterwards by telling the model which of several outputs is the best. there's obviously a lot of finer details where innovation happens, but it's not anything fundamentally new. the end product is very impressive of course. my measure for innovation is more from the theory side -- did they invent something really novel? and google did a lot more on that front than openAI.
Jump in the discussion.
No email address required.
What are some non-theoretical things that Google has done? As far as I can tell the actual real world applications of this stuff are sending me more targeted ads/extracting greater surplus in their ad auctions.
Jump in the discussion.
No email address required.
you can find various applications on the deepmind website, e.g. the blog.
the research they publish is not all the research they do, and if it's published other teams quickly build upon it and improve or adapt it to various situations. for example AFAIK alphafold is nowadays not the best protein folding algorithm for many types of problems, some alternatives are faster for certain problems, some are slower but for certain problems more likely to find a solution, etc.. For large language models google research used to focus more on efficiency (small model size and fast computation for a given quality) instead of best quality, based on the philosophy that eventually everyone has to start focusing on efficiency anyway.
Jump in the discussion.
No email address required.
More options
Context
More options
Context
More options
Context