Tomas Gear

I'm generally bullish on dictation tools overall, but the efficiency gains follow a clear pareto curve. You're often better off pausing to compose your thoughts than rambling for five minutes to get them across. Speaking also burns more energy than typing, which can easily offset any time saved.

You can self-reflect more often and more vividly with AI because you have an always available, non-judgmental thought partner.

@nic_carter agreed, but it also disproportionately rewards generalists who are dangerous enough across a few domains and have the taste to steer productively

It should be pretty obvious at this point that AI is a "force multiplier" not a "labor substitute". It helps experts be better at things they are already good at. It doesn't let beginners match experts. If you can't write, anything you write with AI will be unmitigated slop. If you aren't a software engineer, anything you vibecode with AI will have security holes and won't be able to scale past a toy demo. If you blindly trust AI to deliver on a research task without knowing the subject matter, you won't be able to fact-check it. There's this weird misconception of AI as something that completely levels the playing field. I don't see it that way at all. There are mathematicians deriving novel lemmas with off-the-shelf models. Normal people can't do that. AI is a tool that makes experts better. It doesn't make everyone into an expert.

Today's most important skill is metacognition

@tbpn @joespeez Welcome @joespeez! Excited to have you here!

BREAKING: Llama lead @joespeez is joining Reflection as VP of Product & Head of Open Source

@reflection_ai @joespeez Welcome @joespeez! Excited to have you here!

Welcome @joespeez to the Reflection team, it's great to have you as our VP of Product & Head of Open Source.

A lot of AI systems progress can be reduced to: partitioning, of some kind proximity, of some kind concurrency, of some kind

There will be human conversations with AI systems that will be so extraordinary they will need to be vaulted or locked as important for humanity, perhaps as inflection points or indications of what is newly possible. If it were to exist, I'd say the vault should be browsable and publicly accessible. Surprised something of the sort doesn't yet exist.

Outstanding panel at GTC on why the future of models is open. Bravo, @MishaLaskin @mntruell @AravSrinivas @miramurati @hwchase17 moderated by the one and only Jensen @nvidia @NVIDIAGTC

The key to good writing is detachment from its first form. Which is a good life lesson overall.

@reflection_ai Part two of the announcements of today - the Nvidia Nemotron Coalition unites all global AI labs that drive open models. nvidianews.nvidia.com/news/nvidia-la…

Excited about this announcement and what it means for @reflection_ai, open intelligence, and the U.S. ecosystem. If you'd like to help deploy frontier AI at scale, DM me. We're hiring FDEs.

Building open intelligence matters because it democratizes the model weights for all and they can be inspected as well as customized independently. Today Reflection is proud to share that its partnership with Shinsegae Group to build a 250-megawatt AI factory for South Korea; sovereign AI infrastructure that Korean government and enterprises can own, control, and build on. This partnership is that belief made real. Excited for what's ahead.

Reflection is partnering with Shinsegae Group to build a 250-megawatt sovereign AI factory for the Republic of Korea. Open intelligence. Built on trust between allies. Owned by the nations that need it most. The future of sovereign AI. Read more in the @WSJ.