1. X
  2. Alex Ratner
Log inSign up
Alex Ratner
Snorkel AI
1,910 posts
Image
user avatar
Alex Ratner
Snorkel AI
@ajratner
@SnorkelAI @uwcse / prev @StanfordAILab – Interested in data management systems for machine learning, weak supervision, and impactful applications.
Menlo Park, CA
ajratner.github.io
Joined November 2013
709
Following
6,850
Followers
RepliesRepliesArticlesArticlesMediaMedia
  • Pinned
    user avatar
    Alex Ratner
    Snorkel AI
    @ajratner
    Feb 15
    This week we launched the Open Benchmarks Grant with a $3M initial commitment from @SnorkelAI + partner support from @huggingface @togethercompute @PrimeIntellect @PyTorch @harborframework & others, in order to close the evaluation gap in AI. Our ability to measure AI has been
    Image
    Open Benchmarks Grant for Agentic AI | Snorkel AI
    From benchmarks.snorkel.ai
  • user avatar
    Alex Ratner
    Snorkel AI
    @ajratner
    Aug 2
    We are once again witnessing a new version of Moravec's paradox, where what we think of as hard vs. easy for humans turns out to be a poor guide for what's hard vs. easy for machines. This is not to express pessimism about the latest Math/Coding breakthroughs (mind blowing!!),
  • user avatar
    Alex Ratner
    Snorkel AI
    @ajratner
    Jul 28
    One of the main goals in AI has always been to get subject matter expertise into computers as efficiently as possible. Data is the most natural, universal transfer medium for subject matter expertise. But it's never really been about data :)
  • user avatar
    Alex Ratner
    Snorkel AI
    @ajratner
    Jul 26
    All enterprises need some kind of specialized AI to be non-commodity in the AI era - and for that, they need a data flywheel. However: data flywheels are *not* built by just passively collecting user/agent interaction traces and then tuning on these. That is like telling a
  • user avatar
    Alex Ratner
    Snorkel AI
    @ajratner
    Jul 24
    Excited that @SnorkelAI was able to support / partner with 1/3 of the benchmarks listed here via our Open Benchmarks Grants program!
    user avatar
    Claude
    Anthropic
    @claudeai
    Jul 24
    Replying to @claudeai
    On several coding and knowledge work evaluations, Opus 5 is the new state-of-the-art:
    Image

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Advertisement
Advertisement