mvkel

> The model isn't wrong — it's just optimizing for the wrong signal.

sniff sniff I feel like I'm going insane.

Weirdly, LLMs seem to break with these instructions. They simply ignore them, almost as if the pretraining/RL weights are so heavy, no amount of system prompting can override it

Accurate info on what? Most markets have almost no liquidity, outside of sports and crypto

Development isn't hard on ram. Doing what Apple claims this device is designed for, spending lots of time in multiple browser tabs, is.

It's strange that the low-end machines get positioned as "every day task" devices when the biggest ram hogs by far are browsers and websites.

It's more that they can't think of anything else that could possibly need that much compute.

> disassembling, evaluating, and feeding back ... listened, iterated, and shipped. they didn’t declare victory and go home. They kept pushing.

Not even an attempt to clear the ai smell out of this piece.

I mean, given your very specific filter, how many companies exist today that were formed in the last decade would be on the list? Stripe?

It's more like "Tesla is going to be bankrupt imminently" all over again.

Name a time in history when a company with an industry-defining product with huge demand failed under the weight of cash flow, or regulation.

Frontier AI is very close to a zero sum game. Focus on profit, and you will lose.

Profit is what you do when you're not trying to grow. Name a time in history when a company with an industry-defining product with huge demand failed under the weight of cash flow, or regulation.

Anthropic and OpenAI are shipping AI agent SaaS and generating tens of billions in revenue. It's safe to say yes.

Excited to try this. Is this not in effect a kind of "pre-compaction," deciding ahead of time what's relevant? Are there edge cases where it is unaware of, say, a utility function that it coincidentally picks up when it just dumps everything?

It's more that I'm allergic to hypocrisy.

Again, these are just words, and they're a salad of legal terms of art that provide cover for virtually any action. "We prevent the -illegitimate- use..." [define illegitimate plz?]

In your second link, that team was defunded; the person heading it just left ceremoniously: https://x.com/mrinanksharma/status/2020881722003583421?s=46

That's my point. They formed anthropic under the sole mandate of "guardrails first," now seemingly don't have them at all. So they're just another ai company with different marketing, not the purely altruistic outfit they want everyone to believe

for sure. If they weren't so self-righteous about not serving ads, it'd be a great revenue stream for them. It'd also align with Dario's seeming obsession with profitability

The "safeguards" you are referring to are contractual, i.e. words. There are no technical safeguards, per the article.

The memo literally says that the reason they have these policies is -because- actual technical guardrails are not reliable enough.

Are the guardrails not part of their core? Isn't that the whole premise of their existence?

Should probably look up how many private companies are suing the government at any one time because of a breach of contract. And that's publicly breaching.

NSA and other three-letter agencies happily do it under cloak and dagger.

"as an ai safety company, we only believe in -partially- autonomous weaponry"

Ads are coming.

These are literally words. The DoW could still easily exploit these platforms, and nothing Anthropic has done can prevent it, other than saying (publicly), "we disagree."

Good optics, but ultimately fruitless.

If preventing mass surveillance or fully autonomous weaponry is a -policy- choice and not a technical impossibility, this just opens the door for the department of war to exploit backdoors, and anthropic (or any ai company) can in good conscience say "Our systems were unknowingly used for mass surveillance," allowing them to save face.

The only solution is to make it technically -impossible- to apply AI in these ways, much like Apple has done. They can't be forced to compel with any government, because they don't have the keys.

Convenience trumps everything, including privacy and security.

Tell the average person that they have to install their own model is a deal breaker at the outset.

As for 99% capabilities being on device, battery life makes it a non starter.

Regardless of their intent, this is the opening move in a ratchet that only turns one direction.

Platforms almost always begin with "90% of users will never notice," then gradually expand scope as regulatory pressure, liability, and competition increase. We saw similar dynamics with real-name policies at Facebook: initially narrow safeguards become infrastructure.

Auditability, fraud prevention, enforcement reporting, etc, historically pushes platforms toward more persistent verification regimes.

"On-device only" promises often erode once regulatory audits demand server-side attestations.

This is the only thing these days that Apple gets right: they make it technically impossible to backdoor, not just promise that they won't.

The second-order effect is normalization. Once large platforms operationalize age assurance, regulators point to them as proof that stricter mandates are "feasible," accelerating a global compliance cascade. Smaller platforms then adopt similar systems to avoid liability.

In past cycles (COPPA, GDPR, cookie banners, payment KYC) the burden disproportionately favored incumbents who could afford compliance. The likely long-term equilibrium is a stratified internet where meaningful participation in adult spaces increasingly requires some portable proof-of-age token, whether nominally anonymous or not.

It's not that Discord is acting in bad faith, they are simply the first domino to fall in what will be yet another GDPR cookie banner, a further erosion of privacy, and another nail in the coffin of a free internet.

Discord should have just "poasted through it," Huberman-style, and the mob would move on to the next platform that will inevitably be forced to enact similar policies.

Good question! It serves as a bootstrap for each module design file (effectively a deeper expansion of what is defined in the PRD). Once you have (at least) design files for each module, PRD is no longer necessary. It usually ends up being inaccurate after the brainstorming phases, which change fundamental bits of the wiring that the initial brainstorm wouldn't cover.

Excited for the inevitable "Windows 3.11 - 98K MRR" in pieter's X bio.

From Claude docs: Planning is most useful when you’re uncertain about the approach, when the change modifies multiple files, or when you’re unfamiliar with the code being modified. If this isn't true, skip the plan.

Love this forum UI/UX. It's clear, snappy, intuitive. How far we have fallen.

That said, allowing posts with no auth is a choice.

> This isn't another trading bot. It's an autonomous agent with memory, intuition, and the authority to act.

The AI smell sure is strong with this one.

Overall, this bot will fail to generate a profit; it is trading what is already confirmed, using data that everyone has, ergo there is no edge.

> the workflow I’ve settled into is radically different from what most people do with AI coding tools

This looks exactly like what anthropic recommends as the best practice for using Claude Code. Textbook.

It also exposes a major downside of this approach: if you don't plan perfectly, you'll have to start over from scratch if anything goes wrong.

I've found a much better approach in doing a design -> plan -> execute in batches, where the plan is no more than 1,500 lines, used as a proxy for complexity.

My 30,000 LOC app has about 100,000 lines of plan behind it. Can't build something that big as a one-shot.