building đź‘ľ https://xitter.cf/t.co/fuKpzUIJ8E | cooking @israelgov
Joined May 2019
- Tweets1.2K
- Following524
- Followers144
- Likes1.8K
Vercel just did to agents what Next.js did to the web. They released eve: an open-source framework where an agent is just a directory of files.
agent.ts → the model it runs on
instructions.md → who it is
tools/ → what it can do
skills/ → what it knows
subagents/ → who it delegates to
channels/ → where it lives
schedules/ → when it acts on its own
We launched a free Vercel Academy course on building @eve agents.
Learn how to ship a durable agent to production: set up a project, call tools, add a human-in-the-loop, and deliver the agent through @SlackHQ. vercel.com/academy/building-…
Vercel just did to agents what Next.js did to the web. They released eve: an open-source framework where an agent is just a directory of files.
It handles the model it runs, who is it, what it can do, what it knows, who it delegates to, where it lives, and when it acts in its own.
We launched a free Vercel Academy course on building @eve agents.
Learn how to ship a durable agent to production: set up a project, call tools, add a human-in-the-loop, and deliver the agent through @SlackHQ. vercel.com/academy/building-…
Genuine question: if coding got so easy and everyone is a genius now why is everyone building useless apps instead of trading algorithms that would actually make them fortune?
Maybe, just maybe, the hard part of software was and still is thinking.
Half my feed: SaaS is dead, software stocks crashing, etc.
Other half: Spinning up AI agents in parallel to build apps non-stop.
But wait… if software is buried who's going to use or pay for this endless stream of new apps everyone's racing to build?
2026 is wild.
We are going to have some fun with this 1M token context
Seems interesting to check. Safe zone for multiple agents at once? Who tried it already?
We’re excited to launch the Codex app, a command center for building with agents.
It gives you a focused space to manage multiple agents at once, run work in parallel, and collaborate with agents over long-running tasks.
openai.com/codex
This insane!
Larry and Sergey can’t stay in California since the wealth tax as written would confiscate 50% of their Alphabet shares.
Each own ~3% of Alphabet's stock, worth about $120 billion each at today's ~$4 trillion market cap.
But because their shares have 10x voting power, the SEIU-UHW California billionaire tax would treat them as owning 30% of Alphabet (3% Ă— 10 = 30%). That means each founder's taxable wealth would be $1.2 trillion.
A 5% wealth tax on $1.2 trillion = $60 billion tax bill, each.
That's 50% of their actual Alphabet holdings—wiped out by a "5%" tax.
Section 50303(c)(3)(C) of the 2026 Billionaire Tax Act states: "For any interests that confer voting or other direct control rights, the percentage of the business entity owned by the taxpayer shall be presumed to be not less than the taxpayer's percentage of the overall voting or other direct control rights."
This means if a founder holds shares representing only 3% of economic interest but 30% of voting control (through Class B supervoting shares), the tax would presume their ownership stake is at least 30% for valuation purposes, not 3%.
The wealth tax is poorly defined and designed to drive tech innovation out of California.
AI is an amplifier, not a savior:
> Low performers + AI = Faster technical debt.
> High performers + AI = 10x output.
Calculators didn't turn us all into mathematicians. The hard part was never the syntax it was knowing what to solve. The illusion of competence is dangerous.
In early stage products, founders will go out to look for more traffic to get more users, but the right thing to do is to focus on turning the early adopters into promoters.
Fail enough and you’ll find success. If it’s not hard, it’s not worthwhile as above, so below.
