Ben House

@waynesutton @convex Gmail and yes dark mode

@beninthehouse @convex Hi sorry about that, are you in darkmode? what email client?

@convex I love your putting on a conference, but I can’t read any of this, my eyesight is already cooked enough 😭😭😭

@theo I love you vision and how you think about problems. Building off what you see in the coding market and adapting for marketers has been the single best strategy for my business lol.

T3 Code is one of the best agentic code tools right now. I genuinely believe we're about to leap frog the others with this next update. The "end to end" vision is so close to realized. I'm so hyped to share this with you all. I'm proud that we get to do it entirely in the open.

@shengkun_ye Ready to test out and see if our users enjoy!

We just killed Apify. Your agent can now read every social media platform. No logins. No subscriptions. X, Reddit, LinkedIn, TikTok, Facebook, Instagram, YouTube, Rednote, and even Amazon. Apify: $199/month. Monid: From $0.0015/request. Pay as you go.

@shengkun_ye Very interesting will test out on daoco, I’m not super keen on apify so built a bunch myself.

Try it now: monid.ai/blog/social-me…

@shadcn Creativity value goes 📈📈📈

What happens to creativity when AI makes copying effectively free?

@jamwt Been building the solution to this exact problem at daoco. We are 1 feature away from feature parity or better. Built entirely on convex :)

I'm compelled to use claude code for content right now, and man, opencode mogs it so hard. This is just not an impressive or polished TUI app.

ChatGPT vs daoco - who actually knows your brand? We gave both the same brand (Spotify) and one prompt. daoco imported the site, knew the colors and voice, and generated a "New Music Friday" social asset instantly. Try free at daoco.org

I have been back in the US for more than 24 hours and my session limits are still on Lisbon time.... Claude what do I do?

@claudeai @trq212 Bring me another!

We're extending Claude Fable 5 access on all paid plans, as well as keeping Claude Code’s weekly rate limits 50% higher, through July 19.

I honestly do quite often, mostly for just executing an already defined specific task. For example If the model already has the context/plan for what I want it to do. I use low to execute and find it’s much quicker, token efficient, and less likely to go off in the wrong direction.

Who’s out here running SOTA models on ‘low’?

Makes sense, coding harnesses set way too high of a bar for technical expertise even though it mostly does everything for you. I think beyond the generic “work” tools there is a real opportunity for even more domain specific harnesses. These are too specific for massive businesses like OpenAI/Anthropic/Cursor, but could still easily be multi billion dollar markets. Trying to build on for marketing with daoco.

Cursor is developing a general-purpose AI agent called Sand as it looks beyond coding tools for its next wave of growth. The product could put Cursor more directly against Anthropic’s Claude Cowork and OpenAI’s ChatGPT Work. More details: thein.fo/4eYkBA8

@jjfleagle 100%. The models are finally getting reliable enough. But the real challenge is the harness around them: context, tools, workflows, and UX that let them actually do useful work. That’s what I’ve been building with daoco.

@beninthehouse The feature that matters is the one that removes an entire handoff. If AI drafts an answer but a person must still find the files, rebuild the deck, update the tracker, and send the follow-up, the workflow is only partially automated.

@SovietInvestor @thekaransinghal Might be as close as the US gets to universal healthcare lowkey

@beninthehouse @thekaransinghal No. They like keeping us on monthly subscription services. That's a big one they plan on charging us, forever.

♥️ GPT-5.6 is a major step forward for health, both at the frontier and at cost. These models push the frontier of performance per dollar, bringing the best health intelligence to all. The smallest variant, GPT-5.6 Luna, evaluated at the lowest reasoning effort, outperforms GPT-5.5 at the highest reasoning effort–despite costing 25x less. The largest variant, GPT-5.6 Sol, sets a new high bar at cost. Another especially cool result: physicians found fewer flaws in GPT-5.6 responses than physician-written responses. We collected diverse tasks that remain difficult for recent OpenAI models, across patient-facing and clinician-facing use cases. We asked speciality-matched physicians to write responses to these tasks with unlimited time and web access. We then asked other physicians to compare responses side-by-side, blinded to their source. Physicians were asked to comment on areas of improvement across five axes: accuracy, communication, completeness, instruction following, and health decision helpfulness. We then reported the fraction of responses across sources rated perfectly across all axes, across 20,000 total axis ratings. GPT-5.6 Sol appeared strongest, although all GPT-5.6 models performed significantly better than physicians.

So interesting because I grew up roughly knowing the other cofounder Sophia Kianni who was a hard working and smart individual focused on climate change. This article never mentions her anywhere, just the engagement baiting the Gates name. To be clear not defending any foul play, but its interesting and probably a side affect of VC expectations IMO.

A startup co-founded by Phoebe Gates that bills itself as a “personal shopping assistant” took credit for online sales it didn’t actually drive, a Bloomberg investigation found bloomberg.com/news/articles/…

Next up in the brand onboarding series @LiquidDeath a world renound brand known for its unique branding, but daoco can easily onboard in less than 60 seconds even with a pesky SVG logo on their site!