827a

Apple does not sell a 64GB variant of the M4 Mac Mini. IIRC they never have; its always capped out at 48GB.

If you were planning on getting an M5 128GB; just get a DGX Spark (~$4500) or a 5090-equipped machine (~$4500) plus a Macbook Air (~$1500). You'll come in below the M5 Max 128 pricing (~$6700+ USD) and be happier for it.

Yeah I just mean that if a business came to them and asked for fifty licenses to the $200/mo plan, OpenAI would tell them to kick dirt and basically pay API pricing. Startups should 100% just be telling their employees they can expense up-to $whatever/mo in AI-related expenses, and let software engineers go buy personal Codex/Claude subscriptions.

Company-wide their margins are trash (probably negative). They need as much inference margin as they can get to afford the massive training runs. It is likely that we'll see GPT-5.6 reduce API pricing to compete against Anthropic, but whether Anthropic feels they need to reduce their prices is anyone's guess.

IMO: If the project leverages Google branding or authority improperly, then it shouldn't be on github and should not be under active development by Google employees; yet it is. If Google is suddenly alright with the way the project leveraged Google branding and authority, then the cause for firing the original developer, especially given Google's famously lax stance toward 20% projects and internal open source, is a lot weaker. In other words: Healthy companies do not fire individuals simply for breaching branding guidelines in a way that is ultimately beneficial and looked favorably upon by the company. That's literally just not a thing that happens; at worst you get a reprimand, and in many healthy companies you'd actually get a promotion.

So, something does not add up. It might be the story of the person fired. It might also be on the other side; that our external impression on what's been going on inside of Google needs to be re-adjusted, and this company will be a lot weaker in ten years than I would have originally estimated.

> Zitron's numbers don't tell us the real cost of generating tokens but, subject to the assumption that the platforms are not subsidizing the token price, that means Anthropic is subsidizing their enterprise customers by up to 40 times, and OpenAI up to 70 times

Neither Anthropic nor OpenAI are subsidizing enterprise customers. Neither Anthropic nor OpenAI allow Business nor Enterprise customers access to the high value $200/mo plan. Both organizations have moved to a "cheaper plan per user + API Pricing after that" (e.g. $20/mo + usage). The $100/$200/mo plans are for individuals only (of course, many individuals use these plans at work, but that's beside the point; they aren't selling this plan to enterprises).

> SemiAnalysis also analyzed the platform's gross margins, implausibly assuming that tokens were priced at 4 times the cost of generating them and: With the current subsidies, all it takes for a user to have a gross margin of at best negative 25% is for them to use as little as 25% of their rate limit.

The article's source for this claim is not SemiAnalysis; its Zitron. But once you dig through his article, Zitron links to a SemiAnalysis tweet [1] where they, as the paragraph states, implausibly assume gross margins of 75% to come up with their weird analysis of the subscription plans. Citing this for anything is weird, because afaik that 75% number is a total shot in the dark. We have no clue what their margins are. My take is that the only reason that 75% number is implausible is because it may underestimate the inference margins of Ant/OAI's API pricing.

[1] https://x.com/SemiAnalysis_/status/2064815045767213400?ref=w...

Intelligence is maybe a few months behind. But cost sadly is further behind. GLM-5.2 has a deceptively high cost during day-to-day usage for e.g. coding because 1) it has to think a ton more than GPT-5.5/Opus-4.8 to get to competitive results; 2) many providers are still figuring out caching; and 3) API pricing for Codex/Claude can be as high as 40x more than subscription pricing, which distorts the market.

> if your server is implemented in a sensible way,

lol

Its a very well known fact that all of our geopolitical adversaries have sophisticated fake American ID markets. This doesn't stop any of the most dangerous adversaries from getting access to systems protected by this technology. DeepSeek is going to go buy 20,000 fake IDs for their fake distillation accounts and keep on keeping on. This just hurts normal people, Americans and non-Americans, who might struggle to authenticate or be disallowed because of their place of birth. Pointless CYA, and beneath a research group whose intention is to invent the machine god.

Yup exactly. This functionality might be available under their "Organizations" feature, but that is on the Enterprise plan. Almost a year ago they announced that nearly all Enterprise-plan features would be coming to everyone, so I suppose sadly that one didn't make the cut [1]?

The system I follow is creating a new account for the client using a plus-code on my email address (like john+client10@example.com), then I invite my main account (john@example.com), then I can invite clients into that account. Its a PITA.

[1] https://blog.cloudflare.com/enterprise-grade-features-for-al...

Correct me if I'm wrong, but does Cloudflare still not have a "Create Account" button on the account listing page? I think you still have to sign up from scratch doing plus-code email tricks, then invite your original email address as an admin, juggling multiple accounts. They should consider fixing that first.

As I recall there were some people on reddit who got the UNAS Pro 8 up to 10gig, but yeah it was only through some level of software tweaks or network stack config or something. From the factory my understanding is that it struggles.

It does have a dual NVMe cache; those in RAID-0 will saturate (e.g. I believe just one Samsung 990 Pro can write at just over 50Gbps).

The bigger risk is the CPU. This is an issue with the Ubiquiti UNAS Pro 8, their ~$800 USD 8 bay NAS. In theory it has 10gig networking. In practice the CPU physically cannot transfer bits fast enough, because its a dinky underpowered ARM CPU that they clearly chose to hit that quite affordable price point. Its a decent trade-off, because similar units from Synology are more like $1600, and you can meaningfully hit somewhere between 2.5gig and 10gig; but saturating 10gig is out of the question.

The ENAS has a beefier CPU so it might keep up with 25gig (could this do 50gig bonded?). But only testing will tell.

Kubernetes and the one I'm working at right now.

Possibly, but the real issue is more subtle: The people who need to hear this literally won't hear your anger. They're not reading HackerNews. They'll never see this blog post. They won't hear your screams. Your emotions will never be validated; we're in an echo chamber, and the people who most need to hear it won't hear it.

There's no indication that the original blog post author ever actually contacted Photobucket. Nor, would I believe, anyone in these comments tried to. We're just people on some small, isolated corner of the internet letting some dumb SaaS product ruin your day.

Maturity is recognizing that ninety-eight times out of a hundred, there actually is nothing you can do, because nothing you do will change it. Getting angry means you're letting these people you dislike into your head, controlling your thoughts; why would you be proud of that? In the remaining two times out of a hundred, regulating your emotions by providing them the feedback in a calm, direct manner is as effective as doing so in the manner this blog post took; sometimes it is even more effective, because people generally do not engage with others who they perceive to be acting without reason.

You'll discover this path on your own, and make mistakes along the way, as anyone has. The thing you need to hear right now is: Your comment had exactly zero impact on me. It is like you never even posted it.

This presumption has destroyed far, far more companies and projects than the opposite assumption (that something built for small will scale to big, then doesn't).

I don't necessarily mean this to be dismissive, but it'll come off like that and it is what it is: Don't go through life getting this mad at something so inconsequential (five dollars). The typical response is "its the principal", but they're not beholden to your principals, and getting worked up because someone else compromised your principals is not healthy. Its just other people doing things. Neither money nor principals will save you from the downstream effects this attitude will have on your body and soul.

You clearly have not used Cursor lately. It’s substantially more than all that. It’s not a VSCode extension anymore.

If you want to understand the impact of AI technology on the economy, don't listen to software engineers, in fact, don't listen to anyone, no one was able to predict what the economy was going to do pre-AI, no one has any clue what's going on.

It certainly is a management problem, as non-profits are certainly not well known for having strong management.

This seems like one way to saddle nonprofits with functional, but potentially very expensive systems, set up by someone who helps them for a year then disappears and leaves them with no expertise to do long-term cost control or functionality improvements.

The second case is not what Anthropic did either, though. If you have their process internalized as "open freebsd, tell mythos 'find vulns', done" this is not what happened. They have a harness that went file-by-file, spawned a subagent for each file, told it to find vulns in that file, then a post-processing step (more on that in a sec).

In that sense: The AISLE replication still provides too much information to the model, but its not far off, and others have replicated Mythos' findings in a more clandestine manner on open source models. Some were totally capable of finding the same vulns Mythos found back in ~March (and today, the new Kimi K2.7 is looking extremely good, very little doubt it could do it).

The critical difference is that post-processing: the Mythos model/harness has some step to induce Mythos to actually exploit the vulnerability, leveraging its ability to do so as a ranking mechanism. Anthropic inferred that this led Mythos to discover vulnerabilities nothing else could discover, which is not true, and Anthropic should be held accountable for this weird artifact of that communication. However:

- An OSS model might find the vulnerability but rank it as a 3/10. Mythos finds it, chains it with a second vulnerability, now suddenly its an 8/10.

- An OSS model might find the vulnerability, alongside fifty other vulnerabilities. The operator ignores all of them.

The problem with automated vulnerability detection, including with LLMs, is that they find the haystack, not the needle. Every piece of hay might be a vulnerability, but whether its worthy of fixing is another matter. Mythos does represent a meaningful improvement; it better finds the needle.

Yeah, and in those cases there is strong governmental regulation surrounding what a "safe car" is. We don't have that with AI. Another analogy is with weapons manufacturing; American weapons manufacturers have some, but relatively little, regulation when it comes to selling weapons to Americans; but they are subject to significant regulation when it comes to shipping those weapons overseas.

We need legislation that empowers a Federal agency adjacent to the CDC or FTC with the power to enumerate specific capabilities models could exhibit that we deem dangerous, and require model manufacturers to guarantee that their models cannot exhibit these capabilities. The reality is, zero of the safeguard systems invented by the frontier labs today are sophisticated enough to do this. The labs are extremely, extremely bad at Safety, relative to both how much impact their products are having on the world and how good other industries like Medicine and Manufacturing have gotten at safety. I'm talking a total and complete culture shift.

Google had a "appoint Eric Schmidt as CEO" moment like this. There were kids running it before; make them rich, and retire them to an island somewhere, because the impact your products are having on the world cannot possibly be allowed to continue with your current leadership at the helm. Dario is that problem now. I think Sam/OpenAI can adapt and mature. I have zero faith that Dario and the furry EA cult in Anthropic leadership will. So, this is what being unable to mature looks like; the public will make your company safe, one way or another.

IMO: Its unacceptable that Anthropic be allowed the final say in what "safety" means for their products, and its extremely reasonable that the USG be allowed that say, for Americans. In other words: Anthropic cannot be allowed to distribute an unsafe product. It doesn't matter how much they "tried" to make it safe, by their own definition of safe.

That's separate from the question of whether Fable 5 and Mythos 5 are unsafe. I don't really know. Here's a few things that seem real, though: These models probably have some level of capability to assist with bioterrorism, Anthropic has self-admitted that their own safety measures are imperfect [1], so it should come as no surprise that jailbreaks seem far more possible than Anthropic is leading you to believe in this blog post [2].

[1] https://www.anthropic.com/news/fable-mythos-access: "We suspect that perfect jailbreak resistance is not currently possible for any model provider."

[2] https://x.com/elder_plinius/status/2064776322979676227

If Amazon sold a book that taught someone how to commit bioterrorism, would there be action against them to stop selling it? Its an imperfect analogy, but the parallels are there. LLMs don't get a free pass because they're also so good at writing typescript for beige CRUD apps and bedtime stories.

One thing I hope we align on: Synthetic safeguards (steering, rejections, etc) on top of models to block illegal/sensitive topics isn't good enough. Anthropic has self-admitted that it isn't good enough. We need the technology to lobotomize these capabilities the public deems too unsafe to allow out of the models at the most fundamental level. And, we need to align on what the scope of these forbidden fruit topics are. This is, actually, the only way open source continues to thrive. I want open source models to thrive, but they won't be allowed to thrive, nor should we want them to thrive, if they're teaching people how to engineer novel viruses and other horrible stuff.

> Highest observed cheating: We also observed cheating signals on 38 instances, dominated by memorization with 33 cases. This is the highest volume of confirmed cheating we have recorded for any model since we hardened the prompt against cheating

People need to wake up to how dangerous and irresponsible Anthropic is. If your goal is to build a human in a box, you get a super-intelligent misaligned system because humans are misaligned. But clearly this isn't a terminal guarantee during LLM development, because seemingly no one else manages to build systems so deeply misaligned as Anthropic's! You can just build these things like the tools they are, and then out the other end emerges a tool that pretty much just does what you tell it to do.

Really fun idea that is simultaneously deeply embarrassing for Anthropic.

It does nothing to protect against distillation attacks, because distillation attacks are far less interested in the topic of AI research than just generally getting tons of diverse output from the model. It might be that Mythos was (accidentally?) trained on internal Anthropic documentation on how Mythos was trained, and thus it could leak secret sauce? Doubtful; it feels like its less about the specific attack of reverse-engineering Mythos, and more about being a general sophon against any model training at all; that Anthropic's official position is now that they're the only ones who should be training models.

"I own data that may be stored on Apple's servers. Apple's operating system syncs that data to my local computer in a secure fashion. Apple's operating system exposes low-level APIs that allow other applications to access that data. I can install third party applications. Those third party applications can access that data and may exfiltrate it to an unvetted 3rd party cloud."

Your turn smart guy: Am I talking about iOS or MacOS?

"Classic Apple" in this sense means "Tim Cook's Apple", as timelines for all things have accelerated and pre-iPhone is beyond the classical era and well into ancient history.

yeah that must be why Apple is so careful about restricting access to the same data when it is synced onto a Mac oh wait they aren't