2026-04-06

Why OpenAI doesn't focus on roadmaps from @embirico (Codex product lead):
"At @OpenAI, you either plan near or long-term, but you never plan medium-term. It's just too difficult.
Near-term is up to eight weeks from now. What is a concrete thing you can rally a team to ship? [Long-term] is having a vibe that's like, a year from now on how the models will be smarter.
The product roadmap is an in-between thing that's just kind of awkward. We basically don't have a roadmap.
We have a combination of a long-term direction and short term bets that we think bring us in that direction."
📌 Watch the full interview here: https://youtu.be/9qXc-THAvc0
Peter Yang: "We write very few specs on the Codex team. We're talking 10 bullet points and that's it."
Here's my new episode with @embirico and @romainhuet where they gave me an inside look at how OpenAI's Codex team operates:
→ Live demo: Building in seconds with Spark
→ How the team

91-year-old retired electrical engineer builds a full-scale app for his church using @claudeai + @Replit @amasad
No coding background.
Less than $350 spent.
Proof it’s never too late to build something incredible.
Full video: https://youtu.be/-xDQrJmVcfU?si=KluFa8O-nVWc5OAj

I've wanted a voice AI to let me go through my emails on the way to work for years.
GPT 5.4 just one-shotted it.
https://github.com/tomblomfield/voice-email

SOUL.md - who I am
MEMORY.md - what I've lived
DREAMS.md - what I'm becoming

It’s amazing to watch people figure out how to use this new alien intelligence.
Every couple of days there’s a mini “eureka”
It’s addictive.

Truth wins out, @amasad and @Replit !
Amjad Masad: After 4 years on the App Store, Apple suddenly blocked our updates but our app continues to organically grow and rose to #1 again!

Codex is @OpenAI's fastest growing product today.
Here are my top 5 takeaways from @embirico and @romainhuet on how the Codex team operates:
1. Plan near or long-term, never medium-term.
Near-term is a concrete thing you can rally a team to ship in 8 weeks. Long-term is a directional vision for where models are heading. Medium-term roadmaps are awkward because this space evolves too fast to plan 6-12 months out.
2. The Codex app almost didn't get built.
The IDE extension was already popular and the CLI had momentum. Building the app was "actually quite contentious." The bet was to make the simplest interface to manage multiple AI agents at the same time.
3. Observe what power users do, then simplify it for everyone.
Since the Codex harness is open source, cutting-edge devs fork it and experiment with unreleased features. The team watches how they use it, learns from their patterns, then packages it into something simple for all users.
4. Every role on the Codex team is blurring together.
Designers write more code than engineers did six months ago. Engineers talk to users. PMs send pull requests. Alex himself uses Codex to summarize Slack feedback, post follow-ups to Linear, and explore the codebase. "It's often faster to send a PR [as a PM] than to get someone else to prioritize that task."
5. When hiring, Alex ignores credentials and wants to see what you've built.
When someone DMs him about joining Codex, he looks for a link to something they've built. If it's just a resume, he's much less likely to engage. "I have no idea where applicants went to college. Just show me what you've built."
📌 Watch the full episode here: https://youtu.be/9qXc-THAvc0
Peter Yang: "We write very few specs on the Codex team. We're talking 10 bullet points and that's it."
Here's my new episode with @embirico and @romainhuet where they gave me an inside look at how OpenAI's Codex team operates:
→ Live demo: Building in seconds with Spark
→ How the team

Proud to have been part of this. We outline policy ideas for the transition to superintelligence, to build an open economy where everyone benefits and a society that is resilient to the risks. Progress is fast, and we must navigate these issues urgently. https://openai.com/index/industrial-policy-for-the-intelligence-age/

Proud to have been part of this. We outline policy ideas for the transition to superintelligence, to build an open economy where everyone benefits and a society that is resilient to the risks. Progress is fast, and we must navigate these issues urgently. https://openai.com/index/industrial-policy-for-the-intelligence-age/

I've learned more from working at @every for the past five months than I did in the previous five years of my career. It's an incredible place to live at the bleeding edge of AI, and we're hiring for five new roles on the team.
One small example of what's different here: At previous companies, a custom landing page like our new careers home would have taken 3-6 weeks to launch, requiring coordination across multiple teams and contractors. It was slow and frustrating. We built this in a day using Claude Code with the Figma and Notion MCPs, pulling copy and page architecture from our existing codebase, design system, and internal knowledge hubs. While an agent jammed on the page in the background, I kept pushing on bigger growth bets. We did a quick round of reviews and then shipped it in under 24 hours.
A lot more companies will look like @every in the next couple years. Come join us and show people how to build better businesses and live better lives with AI.

Re Check out all of the roles here: https://every.to/careers
The two jobs on the growth team also have agent-facing applications that get a priority look:
https://github.com/EveryInc/apply-head-of-social
https://github.com/EveryInc/apply-gtm-engineer

Introducing Replit AI SDR
Find your next 100 customers with AI

I met @danshipper @lucas__crespo @katelaurielee months ago. I love what they're building. I would have joined here instantly if I wasn't in convo with Ramp.
They're building the future. I think many late-career tech folks will start small studios/collectives like this, build an audience and their own products. A lot of agencies are stuck with either an advertising model or having clients. Every has earned an audience with sharp storytelling and are now building their own products.
Most of all, these folks are builders at heart. Humble. Passionate. Down to earth. Learners.
I told @lucas__crespo after meeting Dan and Kate that the team gives me new Bauhaus vibes.
Dan Shipper 📧: we just opened up 5 new roles @every:
- GTM engineer
- Head of Finance Vertical, Consulting
- Head of Learning and Development
- Head of Product Marketing
- Head of Social
if you want to help discover and define how the world works with agents over the next 10 years—join us:

Re @Replit to become the BizOS in the AI era.
Samuel Spitz: Introducing Replit AI SDR
Find your next 100 customers with AI

We have 5 new roles open at Every. Probably a 6th soon. Every is an incredible place... it's a playground for builders.
If you think you are one of the best at what you do (or could be), want to work in a truly agent native environment, be on the edge of AI, work alongside cool and caring people, and build more than you talk, you should work here → http://every.to/careers
Open roles:
- GTM engineer
- Head of Finance Vertical, Consulting
- Head of Learning and Development
- Head of Product Marketing
- Head of Social
Dan Shipper 📧: we just opened up 5 new roles @every:
- GTM engineer
- Head of Finance Vertical, Consulting
- Head of Learning and Development
- Head of Product Marketing
- Head of Social
if you want to help discover and define how the world works with agents over the next 10 years—join us:

The future requires new ways of applying for jobs. If you're an engineer and love go-to-market too, apply to Every's GTM Engineer role here: https://github.com/EveryInc/apply-gtm-engineer
I think it'll probably be a top 3 most fun role at Every

(🧵1/11) For the past year and a half, I've been investigating OpenAI and Sam Altman for @NewYorker. With my coauthor @andrewmarantz, I reviewed never-before-disclosed internal memos, obtained 200+ pages of documents related to a close colleague, including extensive private notes, and interviewed more than 100 people.
OpenAI was founded on the premise that A.I. could be the most dangerous invention in human history—and that its C.E.O. would need to be a person of uncommon integrity. We lay out the most detailed account yet of why Altman was ousted out by board members and executives who came to believe he lacked that integrity, and ask: were they right to allege that he couldn't be trusted?
A thread on some of of our findings:

Big things happening at @every this quarter and beyond. Come join the party!
Dan Shipper 📧: we just opened up 5 new roles @every:
- GTM engineer
- Head of Finance Vertical, Consulting
- Head of Learning and Development
- Head of Product Marketing
- Head of Social
if you want to help discover and define how the world works with agents over the next 10 years—join us:

some more ramblings from working at @AnthropicAI.
I've been asked a few times what the single most important thing a growth marketer should be doing with AI that most aren't.
surprise, it's not just a single specific task. after running dozens of growth workflows through Claude, I think the useful stuff worth doing falls along four dimensions 🧵

Go work with @danshipper & @every EOM
Dan Shipper 📧: we just opened up 5 new roles @every:
- GTM engineer
- Head of Finance Vertical, Consulting
- Head of Learning and Development
- Head of Product Marketing
- Head of Social
if you want to help discover and define how the world works with agents over the next 10 years—join us:

1/ I’ve been at @every for more than 3.5 years, and this is the biggest growth moment we’ve seen.
When I started, it was a newsletter. Now it’s a mix of media, product, and consulting.

My favorite little hack in Claude Code is asking its confidence level before letting it ship something. Anything under 90% confidence and I ask it to go find improvements until its more certain.
Without a real technical background, asking a simple question like this dramatically improves everything from growth experiments to product PRs, like fixing this typing indicator bug on Plus Ones at @every.