Remote @t3dotcodes over Tailscale feels like magic. Zero lag. For anyone actively using remote / remote-control on the native apps, would recommend giving it a try. Total game changer for me, can now rely on remote rather than it feeling like a total lottery.
1
96
Suddenly blew through my weekly Codex limit on $200 pro plan overnight. Did leave 1 goal running, but never gotten anywhere the near the limit before (not sure even seen below 90%), to see it wiped overnight seems weird. Anyone else seeing the same? @thsottiaux
1
6
177
Great product from @danshipper and the @every team here, working in what might be the most interesting space in software right now.
The “death of SaaS” and rise of purely agent-driven workflows are in vogue, but the real design space might be in the middle, where humans and agents collaborate together in their preferred interface.
We've seen this in coding, now the net is widening.
BREAKING:
Proof—a new product from @every
It’s a live collaborative document editor where humans and AI agents work together in the same doc. It's fast, free, and open source—available now at proofeditor.ai.
It’s built from the ground up for the kinds of documents agents are increasingly writing: bug reports, PRDs, implementation plans, research briefs, copy audits, strategy docs, memos, and proposals.
Why Proof?
When everyone on your team is working with agents, there's suddenly a ton of AI-generated text flying around—planning docs, strategy memos, session recaps. But the current process for collaborating and iterating on agent-generated writing is…weirdly primitive.
It mostly takes place in Markdown files on your laptop, which makes it reminiscent of document editing in 1999.
Proof lets you leave .md files behind.
What makes Proof different?
- Proof is agent-native: Anything you can do in Proof, your agent can do just as easily.
- Proof tracks provenance: A colored rail on the left side of every document tracks who wrote what. Green means human, Purple means AI.
- Proof is login-free and open source: This is because we want Proof to be your agent's favorite document editor.
Check it out now, for free—no login required:
proofeditor.ai
1
1
151
Paul D retweeted
Lesson in that.
Hasbro is being sued by its own shareholders for printing so many damn Magic cards, 'destroying the long-term value of the brand' pcgamer.com/games/card-games…
171
323
5,080
390,315
Paul D retweeted
My read on the bill was v different so can someone smarter help me understand...
+ the bill doesn't ban tokenized equities, just regulates them like securities. Limits what you can do on-chain bc of AML/KYC rules but that seems pretty normal, no?
+ there does seem to be some grey area around DeFi, but if you're a non-decentralized app that follows basic operating rules like treasury sanctions you should be ok and don't need AML/KYC on all your users. I expect this to get more clarity over time and litigated through real-world implementation
+ Leaves it to the SEC to determine what is a security, but then regulation goes to the CFTC. I don't think this erodes the CFTC's authority...
+ The bill allows for rewards on deposits, but doesn't call it yield because that would open up a can of worms related to banking laws. From a technical standpoint, yield on USDC and other stablecoins operates more as "rewards" on USD deposits elsewhere, so I don't see what the issue is with this language since it defacto provides users return on holdings
Would love some help understanding the crux of these issues, as they are very important!
@brian_armstrong @SenatorTimScott @SenLummis @NeerajKA
After reviewing the Senate Banking draft text over the last 48hrs, Coinbase unfortunately can’t support the bill as written.
There are too many issues, including:
- A defacto ban on tokenized equities
- DeFi prohibitions, giving the government unlimited access to your financial records and removing your right to privacy
- Erosion of the CFTC’s authority, stifling innovation and making it subservient to the SEC
- Draft amendments that would kill rewards on stablecoins, allowing banks to ban their competition
We appreciate all the hard work by members of the Senate to reach a bi-partisan outcome, but this version would be materially worse than the current status quo. We’d rather have no bill than a bad bill. Hopefully we can all get to a better draft.
We'll keep fighting for all Americans and for economic freedom. Crypto needs to be treated on a level playing field with the rest of financial services so we can build this industry in a safe and trusted way in America.
3
2
6
846
Paul D retweeted
If you think you CAN'T learn something, you're right
If you think you CAN learn something, you're also right
3
1
12
2,251
Challenging to accept, but @DrPippaM's analysis has been consistently prescient. Thought provoking and worth your time.
American Glasnost, Stockholm Syndrome, Crossfire Cyclone and Missing Bitcoins? Link in replies.
1
97
"Figure out, ok, what’s the totally new form factor? And how do I be worse on all the checkbox dimensions, but so good on the thing that matters." - @danshipper
A single quote clarified the pivot we needed to make.
1
80
Another awesome Vibe Check from @every. Just the right amount of information.
NEW!
Anthropic just dropped Claude 4 Sonnet with a 1 million token context.
We've been testing it thoroughly for a few days internally @every and here's what we found:
Long context text analysis:
We hid two movie scenes in 1 million tokens of context, and asked Claude to find those scenes and do a detailed analysis of them in one shot.
It performed faster than Gemini 2.5 Pro and Flash with fewer hallucinations. However, its responses were less detailed.
Long context code analysis:
We loaded the entire codebase for @Every’s content management system (plus some padding to get to 1 million tokens) and asked Claude to do four detailed code analysis tasks in one shot.
Again it performed more quickly than Gemini 2.5 Pro and Flash—but it scored lower than them on this task. It missed key details about control flow and database inheritance.
AI Diplomacy:
We played Claude Sonnet 4 in AI Diplomacy to see how it would perform at world domination.
Claude 4 was surprisingly good at this. Normally, Claude gets taken advantage of in Diplomacy because they don’t often lie. With aggressive prompts Claude Sonnet 4 came in second only behind o3.
1
83
Taking the time to learn @linear finally and annoyed I didn't do it sooner. It's so much better than Notion for managing Product workflows.
One feature request: Ability to create a draft Github PR & Branch from a Linear Issue within Linear, with the issue content included - Awesome for having agents handle small tasks when you're AFK.
5
228