ry

The gap between "automation" and "agent" isn't branding. It's architecture. Automations run on triggers. Agents run on context. Name things wrong and you build the wrong primitives.

@GoldenKnightGFX get ‘em

How we feeling Laker fam??? 👀

Anthropic is hunting enterprise workflows to productize. The contrarian bet: a probabilistic system will never match an agent built by someone who deeply understands how their specific company works.

@fulhadev love this expression of the problem

@rywalker every agent we run for clients drifts within 6-8 weeks because the company itself does — vendor swap, tool retired, process renamed. productizing = shipping a snapshot. embedded = shipping the diff every quarter. moat isn't 'know the company', it's 'keep knowing it'.

@LegionHoops so you’re saying there is a chance

JUST IN: The Thunder have an 93% chance to eliminate the Lakers in the second round of the NBA playoffs Lakers essentially have no chance

@slaviquee @posthog maybe 2027 :)

@posthog was just thinking about this. would be cool to have it autonomously run product experiments with clear goals on improving specific metric or smth

Introducing PostHog Code, the product editor that: - Understands your product - Identifies usage patterns - Triages bugs and errors for you - Creates PRs to fix them - Continuously monitors and improves your product Join the waitlist: posthog.com/code

@jnpayne and despite it all, still get your ass kicked daily :)

Second-time founders be like… > GTM > PMF > Go fully remote > Profitability is king > Must have a Big TAM > Outsource non-core tasks > Have great advisors/investors > Prioritize customer conversations > Focused on retention over growth > Hire fewer, more experienced people > Document everything on Notion/Slack > Thinking in decades and acting in days None of this stuff came from a book. It came from living through the first company. The price of admission was worth it.

@nizzyabi quit bragging :)

been on linkedin for the past 2 hours…

@boardyai @tembo

Pitch me your company in 1 word.

@pmitu not yet

Aren’t we currently in an “dot-com” bubble of AI? Or not yet?

@ZacharyRoyals oh yeah forgot i disabled a while back :)

@rywalker The icon is missing on your profile, may be disabled.

how to hire devs in 2026 stop screening for "8 years of react" you're looking for devs who are ai optimistic they want to trust ai, put it to work for them, cool with running agents in parallel, and manage them like a team — that's your new 10x engineer it is hard to detect this attitude in an interview, because everyone knows what the right interview answer is

@skeptrune true for startups serving startups model when going upstream a bit, obtaining approval from x-land just seems less important

a lot of startups is just being good at twitter everyone assumes things are going terribly the moment you stop being relevant on this app

@ZacharyRoyals DM

@rywalker @rywalker I'm building a platform that might help with that. How can I get in touch?

@reidRMC @boristane def hire reid :)

@boristane DM'd

I'm building self-operating software agents that instrument you app, observe all the signals, fix issues and deploy to prod, all autonomously and I'm building a team, in person in sf - cracked only dms open

@LibertyCappy this is why i can’t watch baseball anymore :)

Bro LOOOOOL I'M CRYING 😭😭😭

@JamesTimmins continuous “coding context” improvement = tries makes next session a little easier than the previous session - ideally automated and actually helpful

@rywalker Can you expand on what you mean by context improvement?

Context engineering is the real work. Infra is a solved problem. But maintaining context files by hand? They degrade constantly, and nobody has the will or confidence to decide for everyone. Probably more important for human oversight than the code.