Log inOpen app
user avatar
David
@dzhng
founder @duetchat • prev: @socialpluscorp @nasajpl
San Francisco, CA
aomni.com
Joined June 2011
699
Following
13K
Followers
RepliesArticlesMedia
  • user avatar
    David
    @dzhng
    Aug 5
    Yes, and it's mainly because all AI agents are all vitamin products. Openclaw, Hermes, Chatgpt, Claude... etc are all vitamins. Only vertical AI products are pain killers. The "agentic" part of AI just doesn't resonate with anyone besides productivity optimizers
    user avatar
    Josh Miller
    The Browser Company
    @joshm
    Aug 4
    Hot take… isn’t it kinda crazy that nobody is really using AI Agents? I don’t mean software engineers or AI early adopters. I mean “college friends talking about it in group chat,” the feeling you got when everyone started using Instagram or TikTok. These frontier AI models
  • user avatar
    David
    @dzhng
    Aug 4
    BREAKING: paying people a shit ton of money attracts people who wants to be paid a shit ton of money
    user avatar
    Polymarket
    @Polymarket
    Aug 3
    JUST IN: Anthropic CEO reportedly increasingly concerned that new hires are working there for a paycheck, as opposed to the company’s mission.
  • user avatar
    David
    @dzhng
    Aug 3
    was just doing an investor update and I just realized we're at almost 17k users now @duetchat we learned a lot the past 3 months, I genuinely think our upcoming update have a real shot at making always-on multi-agent teams accessible to folks outside of the tech bubble
  • user avatar
    David
    @dzhng
    Aug 1
    I've only got loop pilled once I started taking recursive upfront planning seriously. This 2 day run took ~5 hrs of planning, but once you have a few queued up you'll finally actually get to the point where there's an agent working 24/7. Obviously need to be running in the cloud
    user avatar
    Jerry Liu
    LlamaIndex 🦙
    @jerryjliu0
    Jul 30
    Yesterday I cohosted a dinner with @dexhorthy with a wonderful group of founders, to talk about agent loops and loop engineering. Some interesting insights: * Most of our group was *not* actively using /loop in Codex/Claude Code * You can build long-running autonomous agent
  • user avatar
    David
    @dzhng
    Jul 23
    Cheap open source models with frontier advisors is a surprisingly effective combo - Kimi with Fable advisors just beat pure Fable in both task completion rate AND cost / task Just tested this in SWE-Bench multilingual with a subset of 50 tasks. GLM-5.2+Kimi is also a great bang