choppaface

I mean sure but I’m curious to what extent and why it’s so obvious to some. I haven’t seen this sort of language in a marketing post before.

“self-evolves through feedback loops”

Does this mean they distilled Claude? Sounds like what Claude Code will often do.

I wonder if we can trust that it will actually be deleted though based upon what happened to DOGE and Social Security data

https://www.npr.org/2026/01/23/nx-s1-5684185/doge-data-socia...

or maybe they don’t actually cache (fully) but lie and just don’t charge the user right now. at least half the users, who are probably also using the most similar tokens / prompts, wouldn’t really know the difference in latency (or care)

i meant the credit card number itself

credit card itself can’t be used to ad targeting. But plenty of proxies make this point largely irrelevant.

besides proof-of-work, is there any realistic alternative to fingerprinting?

Because Jane Street is unabashed about greed, and yet they do not let that greed lead them into attempting to pollute and derive us of our attention every day.

Isn’t this sort of just “chain of thought” (i.e. the seminal https://arxiv.org/abs/2201.11903 ) where the user is helping the model 1-shot or k-shot the solution instead of 0-shot? I’ve used a similar technique to great effect. I feel things are so new / moving so fast that it’s hard to have common lingo. So very helpful to have a blog / example! But I wonder if the phenomena has been seen / understood before and just in smaller circles / different name.

While I like that you debunked the article . . . I want to hear an argument for where the SWE job market can grow in a post-Claude world. I might expect something like: “CEOs are naturally greedy. So after trimming the team, they then recognized (versus “replacing” people with AI) they could actually accomplish _more_ with more engineers, each empowered with AI.

But I do like folks calling out the OP for being AI spam.

It’s not cutting corners. Apple does most of their testing using strictly internal resources, like secret “mini malls” in the Silicon Valley area. They fail because this testing biases their sampling; users must sign draconian NDAs to participate, among other things. These samples are effectively biased due to Apple’s corporate culture regarding secrecy and competition. So, Apple actually works very hard. It’s just they culturally prefer a lot of techniques that their competitors (e.g. Google and Facebook) have throughly proven as inferior.

But is Google better? Not really, they killed a lot of good products like Reader.

But is Facebook better? Not really, Cambridge Analytica and Metaverse and .. facebook products are disposable.

But I think these Apple UX bugs are misdiagnosed. Yes they are atrocious. But think about how atrocious and non-representative and non-competitive Apple’s testing population is.

Appreciate the nature and scale of the internet... and also how it's changing though, yeah?

While I agree with much of the article's thesis, it sadly appears to ignore the current impact of LLMs ...

> it’s never been easier to read new ideas, experiment with ideas, and build upon & grow those ideas with other strong thinkers on the web, owning that content all along.

But, "ownership" ? Today if you publish a blog, you don't really own the content at all. An LLM will come scrape the site and regenerate a copyright-free version to the majority of eyeballs who might otherwise land on your page. Without major changes to Fair Use, posting a blog is (now more than ever) a release of your rights to your content.

I believe a missing component here might be DRM for common bloggers. Most of the model of the "old" web envisions a system that is moving copies of content-- typically verbatim copies-- from machine to machine. But in the era of generative AI, there's the chance that the majority of content that reaches the reader is never a verbatim copy of the original.

Yes but two other considerations:

1) Assume the buyer/seller holds capital from sources that the majority of the market considers “illicit” and/or is legally sanctioned and/or physically frozen or restricted. Aka the capital can never be called (or at a discount that is unknowable) or the transaction could be later legally reversed or nullified by one or more legal entities. But of course the StableCoin market maker fails to communicate this risk. Therefore the real value of either side of the trade could be zero despite the non-zero StableCoins being transferred. Thus that’s not really a “trade” because there are hidden substantial risks.

2) Along the lines of Matt Levine “Stablecoin treasury strategy?” Consider that the buyer is a publicly listed company, and they fundraise based upon purchase of the digital asset. Then you are doing what most banks consider is not trading but fueling speculation (and normally you can’t expose average retail investors to these risks).

The innovation of StableCoins is much less about Capitalism and much more about re-packaging fraud. And given how lax the prosecution of fraud was during the Financial Crisis, there’s a big meta-bet that StableCoin “traders” will never face losses.

uv can also run even a beefy linux desktop out of file descriptors for larger projects. And does not have deterministic / reproducible installs. Still needs maturity.

A key idea premise is that LLMs will probably replace search engines and re-imagine the online ad economy. So today is a key moment for content creators to re-shape their business model, and that can include copyright law (as much or more as the DMCA change).

Another key point is that you might download a Llama model and implicitly get a ton of copyright-protected content. Versus with a search engine you’re just connected to the source making it available.

And would the LLM deter a full purchase? If the LLM gives you your fill for free, then maybe yes. Or, maybe it’s more like a 30-second preview of a hit single, which converts into a $20 purchase of the full album. Best to sue the LLM provider today and then you can get some color on the actual consumer impact through legal discovery or similar means.

Who has a $2.4m RSU package but only $330k remaining in a mortgage? The four dozen Google L7s who bought foreclosures in East Palo Alto?

A candidate wants a _competitive_ offer. While stock is almost impossible to compare across offers, candidates can at least stack-rank the company’s funding and check to see how the proffered percentage compares to the mean for the funding round. So if a company has high-percentile funding, and gives a high-percentile equity fraction, it’s a good sign to the candidate. But of course, the company could be WeWork, or even OpenAI could get risky if the tender offers stop (which will happen when/if the market crashes).

At the end of the day, it means a lot to the candidate if the company _wants to compete_ for a hire, especially in the current economy (layoff-friendly and SWE saturated, especially versus 10 years ago). A story like “your options could be worth $XXX in 4 years” I hope is not seen as competitive today.

Also know a non-smoker who got lung cancer, in particular a (rather rare) genetically sensitive form. Most of the damage ended up to the bones and brain versus the lungs.

Hugginface is mostly AWS, so these experiments might be done on an AWS-provided cluster? I wonder how many of the results are reproducible on "open market" clusters listed on e.g. https://gpulist.ai/

And/or try contacting a bank (one of the underwriters?) to see if they’ll loan you cash for taxes using the RSUs as collateral. A lot of early Uber employees were able to get loans to exercise and cover taxes, tho these were rather large sums. That said the shares are liquid so less risk.

It’s also important to weight dramatic changes in the White House this week, too, right? There’s the intel itself, but then there’s also the guy trying to control the news cycle.

Need to also consider the ones that had a qualified exit event and then the product got axed (e.g. aquihire or just customer acquisition). It’s a very different graveyard but in many cases has similar impact on the non-Founders (especially the IC SWEs).

And do regularly call in for a discount or fee revision. A few years ago AT&T credited me over $700 for years of over-charging for an improperly disconnected line. I got the refund after about an hour of calling and trying to just get a cheaper plan. You probably don’t want to trust AT&T, but if you pay them remember to treat them as “efficiently” as they do you.

> We cannot find qualified applicants.

By chance has your employer posted on HN who’s hiring?

Can you sell or re-sell colo space to a handful of customers who might put TT or other “weird” hardware there? Doesn’t scale, but it hedges your own business. Requires the right customer though, like somebody who might buy Nervana / Xeon Phi but then buy NVidia from you when it blows up.

> Is what they are shipping valuable?

That’s indeed critical, but most Director-level managers and below have very little control of how well the business model serves the OKRs. Yes the OKRs need to be achieved and help make the business work, but e.g. if the business model’s margins are just too tepid or if the VC’s expected revenue growth (exponential?) will never actually realize, then there is really zero material value to the shipped product. Hence the focus on a happy team that’s shipping, because at least that provides some technological value. And build a network you can bring to your next gig—- because that’s what gets you the next job.

There are rare cases where a team might discover a new business model or impress a whale customer, and then the business model fundamentally changes.

Yes there is risk the “bean counters” or CFO / COO office will want to cut the cord (especially now tech hiring is in a recession). But tech moves fast; those bean counters will likely end up owning shares of a zombie in the next 5-7 years. And their game is to cash out, not build a future.

And if the business model actually works, then keep at those OKRs and everybody should win. Good business models are where stupid can succeed; the team has the right levers.

The main idea behind DSPy is that you can’t modify the weights, but you can perhaps modify the prompts. DSPy’s original primary customer was multi-llm-agent systems where you have a chain / graph of LLM calls (perhaps mostly or all to OpenAI GPT) and you have some metric (perhaps vague) that you want to increase. While the idea may seem a bit weird, there have been various success stories, such as a UoT team winning medical-notes-oriented competition using DSPy https://arxiv.org/html/2404.14544v1

There are a handful of good restaurants (e.g. Sea Pal Cove) and a nice off-leash beach beach there despite it being a very sleepy corner of the small city.

They claim unlimited access, but in practice couldn't a user wrap an API around the app and use it for a service? Or perhaps the client effectively throttles use pretty aggressively?

Interesting to compare this $200 pricing with the recent launch of Amazon Nova, which has not-equivalent-but-impressive performance for 1/10th the cost per million tokens. (Or perhaps OpenAI "shipmas" will include a competing product in the next few days, hence Amazon released early?)

See e.g.: https://mastodon.social/@mhoye/113595564770070726

Discord servers are great, especially if they are tied to an open source project for grounding. Good ones might be hard to find; you might have to try one that isn’t directly related to your interests for a while until somebody there links you to a community or other Discord who has folks you would want to follow.

Branching out like this is critically important even if your company has senior folks who can give you helpful feedback, since your own company will have its own biases.