HN user

greenfish6

118 karma

Building a marketplace for agents. raymond@talkshi.com

Believer in agent to agent commerce.

Tech optimist.

Posts9
Comments72
View on HN

something i really like from tryin git out over the last 10 minutes is that the main agent will continue talking to you while other agents are working, so you don't have to queue a message

GPT-5.3-Codex 6 months ago

yea but i feel like we are over the hill on benchmaxxing, many times a model has beaten anthropic on a specific bench, but the 'feel' is that it is still not as good at coding

i would have to imagine the gastown design isn't optimal though? why 8, and why does there need to multiple hops of agent communications before two arbitrary agents communicate with each other as opposed to single shared filespace?

Excited to try this out. I've seen a lot of working systems on my own computer that share files to talk between different Claude Code agents and I think this could work similarly to that.

(i thought gas town was satire? people in comments here seem to be saying that gas town also had multi-agent file sharing for work tracking)

Claude Sonnet 4.5 10 months ago

pelican on a bicycle benchmark probably getting saturated... especially as it's become a popular way to demonstrate model ability quickly

Claude Sonnet 4.5 10 months ago

As the rate of model improvement appears to slow, the first reactions seem to be getting worse and worse, as it takes more time to assess the model's quality and understand the nuances & subtler improvements

Yes, but if you look in the rate limit notes, the rate limit is 500k tokens / minite for tier 4, which we are on. Given how stingy anthropic has been with rate limit increases, this is for very few people right now

I use claude code in a staging env also with full bash permissions via the open source github actions (I think this is much safer and also what you're supposed to do). The github issues and PR integrations make it really organized

Claude 4 1 year ago

Benchmarks don't tell you as much as the actual coding vibe though

Modern Realty (YC S24) | Founding AI Engineer | San Francisco, CA | Full-time

We're building an AI-powered real estate brokerage that allows users to buy homes without a traditional realtor. Our platform automates agent workflows including messaging, deal management, and customer interaction.

Tech stack: TypeScript, AI/LLM integrations with focus on tool use

Responsibilities: - Design and develop AI-driven workflows for messaging and deal automation - Build robust backend services using TypeScript - Craft effective GPT prompts and fine-tune AI responses - Collaborate closely with CEO/CTO on key platform features

Seeking someone who: - Uses AI to ship code fast - Has shipped consumer-facing features - Has good grasp of existing AI models

Compensation: $100K-$150K, 0.5%-2.5% equity

Apply / JD: https://www.workatastartup.com/jobs/66546

Worked for me once so far for a single github issue which had a solution that involved a read and write to my supabase, seemed to have a big upgrade from the codeact 2.1

Yea definitely playing a bit of persona since comments drive up the rating on HN. 3500 uniques inbound from this post, couple home tours scheduled from inbound users, I'd say I'm happy about it. I would say 75% of people I meet face to face when I say I'm making an ai realtor, it's a negative reaction. Although it does much better in the bay

- Zillow doesn't actually transact homes themselves, once you click that you're interested in a house, a realtor will call you who is non-zillow affiliated, they could be with any brokerage. Redfin only offers you discount services if you buy + sell through them, but most people buying in the bay area aren't simultaneously selling a home through Redfin.

- We are currently using a 3rd party showing service which has people come and literally open doors.

- Yes we are offering representation agreements to help people tour in light of the NAR suit.

- We will solve the blacklist problem when we get to that size. Not going to start running before catching the ball.

- Incumbents like Zillow have agents as their primary customer, they wouldn't put out their main customer. Redfin's model is to have the human agent at the heart of their product, it would have to be completely remade if the customer was using the interface.

- Liens are analyzed using our disclosure summary tool. Solar loans are often liens, so same applies. Inspections are a disclosure. Contingencies fall into 5 buckets which are fairly standard of inspection/mortgage/insurance/title/appraisal. Negotiations are usually done over text message or email.