Log inSign up
john allard
1,113 posts
Image
user avatar
john allard
@john__allard
on sabbatical | prev @openai | watching incredulously from a distance | シットポスト
San Francisco
jhallard.substack.com
Joined July 2020
632
Following
2,628
Followers
  • Pinned
    user avatar
    john allard
    @john__allard
    Mar 21
    I’m taking a few months to ride my bike across the length of Japan, from Cape Sata to Cape Soya. I wrote a post about the route I designed for the trip, and the philosophy, tradeoffs, and process behind building a custom 3,700 km traverse from scratch (link in thread)
    Image
    GIF
    2.5K
  • user avatar
    john allard
    @john__allard
    Nov 20, 2023
    The first OpenAI employee to stop working before 1am. Extremely bearish
    user avatar
    Emmett Shear
    @eshear
    Nov 20, 2023
    Today I got a call inviting me to consider a once-in-a-lifetime opportunity: to become the interim CEO of @OpenAI. After consulting with my family and reflecting on it for just a few hours, I accepted. I had recently resigned from my role as CEO of Twitch due to the birth of my
    153K
  • user avatar
    john allard
    @john__allard
    Apr 22, 2024
    SWEs wandering into the /research directory of jupyter notebooks
    Image
    27K
  • user avatar
    john allard
    @john__allard
    Apr 23, 2024
    Replying to @KenjiCapital
    I would pay much more than $19 and spend much more than 1 extra hour to avoid ever going to LAX
    11K
  • user avatar
    john allard
    @john__allard
    Nov 20, 2023
    Replying to @ashleevance
    Some employees are hard offline (in the wilderness, pat/mat leave), otherwise that list would be basically 100%. Unity has been achieved
    59K
  • user avatar
    john allard
    @john__allard
    Jun 24, 2023
    Replying to @canderaid
    Even better - it was only nominally 5.6% of the budget. A lot of that was the book value of equipment that was sitting in storage and would never be used again, e.g. M113s.
    6.7K
  • user avatar
    john allard
    @john__allard
    Dec 6, 2024
    I really enjoyed giving an early look at our new Reinforcement Fine-Tuning product. The idea that anyone can leverage the same training algorithms and infra we use to create our o1 models and craft expert models in new domains is super exciting. Learn more openai.com/form/rft-resea…
    14K
  • user avatar
    john allard
    @john__allard
    Nov 22, 2023
    Will forever regret not crouching 6 inches lower 🫡
    user avatar
    Greg Brockman
    OpenAI
    @gdb
    Nov 22, 2023
    we are so back
    Image
    37K
  • user avatar
    john allard
    @john__allard
    Nov 22, 2023
    king shit, Emmett - hmu for free api credits
    user avatar
    Emmett Shear
    @eshear
    Nov 22, 2023
    I am deeply pleased by this result, after ~72 very intense hours of work. Coming into OpenAI, I wasn’t sure what the right path would be. This was the pathway that maximized safety alongside doing right by all stakeholders involved. I’m glad to have been a part of the solution.
    37K
  • user avatar
    john allard
    @john__allard
    Sep 20, 2024
    Replying to @Andercot @anduriltech and @SpaceX
    1. this seems insanely destabilizing in the short term 2. don’t hypersonic glide vehicles get around this? we’d end up back in a somewhat less stable equilibrium after a super expensive arms race
    7.8K
  • user avatar
    john allard
    @john__allard
    Jun 24, 2023
    Replying to @canalCCore2 and @fchollet
    the number of folks who have even the vaguest idea that LlaMa exists is entirely negligible compared the volume of search traffic for ChatGPT (and Minecraft). The number who have even tried to run LlaMa applications would be much smaller
    740
  • user avatar
    john allard
    @john__allard
    Aug 12, 2024
    congrats to this great team
    user avatar
    Alistair
    @AlistairPullen
    Aug 12, 2024
    I'm excited to share that we've built the world's most capable AI software engineer, achieving 30.08% on SWE-Bench – ahead of Amazon and Cognition. This model is so much more than a benchmark score: it was trained from the start to think and behave like a human SWE.
    Image
    00:00
    25K
  • user avatar
    john allard
    @john__allard
    Nov 4, 2024
    Wandering the louvre w/ advanced voice mode is a magical experience. See an interesting exhibit, spend a minute practicing my pronunciation of the title/description, get an accurate translation, then jump into a nuanced discussion on the piece and relevant historical context.
    2.8K
  • user avatar
    john allard
    @john__allard
    May 24, 2024
    it’s honestly not that profitable, margin doesn’t mean much when volume is low. real money is and always has been in addictive drugs, it’s why the cartel is moving meth and not mushrooms
    15K

New to X?

Sign up now to get your own personalized timeline!

Create account

By signing up, you agree to the Terms of Service and Privacy Policy, including Cookie Use.

Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Don't miss what's happening
People on X are the first to know.
Log inSign up
Advertisement
Advertisement