The switches with defaults are both switching over a String with only literal cases. That can't be exhaustive without a default, and switch expressions must be exhaustive.
>And in the end, the real issue at the root is that internet traffic, especially egress traffic, still is absurdly expensive, and the behavior of people like Sam Altman that sent RAM, SSD and even HDD prices through the roof is making the situation even worse.
The core problem is that dependencies don't really scale. Find a security issue in one dependency? It has to be updated. All the projects which depended on it need to be updated. Turtles all the way down. AI worsened this further by accelerating the churn with vulnerability scanning. Naturally, they focus on the core projects like apache-commons which fans out to basically everything in the Java universe. Which means maven central becomes overloaded and now needs to make emergency changes to stay solvent.
long-time java dev and did a portion of my startup on dial-up. they've been hostile to any attempt to optimize for bandwidth for a very long time - i had elaborate hoops i jumped through to implement caching that should have been baked in to the tooling
I've been a user of the Java/Maven ecosystem for decades. Maven has had local caching of artifacts for as long as I can remember, and there are several ways to set up a private repository mirror for caching and reproducibility.
The interests of the Maven project (i.e. the Maven community), Sonatype and other actors in the space are not necessarily aligned.
But I do wonder where you had to jump through hoops? A basic Squid caching proxy should be enough at least for a pure pull-through like Maven Central. Once you have to publish stuff, okay, that is messy, you won't get anywhere with just Squid there.
Hey, at least your six year old elementary student hasn't shot you yet... School budgets are perhaps a little thin, but at least you're not struggling to pay out a $10 million lawsuit. So you've got that going for you.
It seems like to me it would be more effective to focus on the object level aspects of their arguments rather than their weird lives? Their argument isn’t “look at my credentials, I’m an expert you must believe me” it’s that “ML research is largely math and science engineering, as AI gets better at math and science it can use that to speed up the cycle at which point something bad could happen”.
These Bay Area people dedicated their lives to AI risk often before AlexNet even showed deep learning could scale, by definition they are eccentric people.
At the same time, people like the AI 2027 guys (who that article speaks poorly of)… have been basically fully on the money so far.
This reminds me of during covid, when one of the biggest public proponents of vaccines in the UK was having an affair. The anti-vaxers then, instead of focusing on whether vaccines were helpful, focused on the affair. That allowed them to dismiss the pro-vaccine point of view without actually engaging on the merits of their argument.
I think the thing is engaging with “ML is mostly math and science engineering, AI is getting better at autonomous math and science, if it gets fully super human the rate of change could be mind boggling” is hard to fully dismiss in the same way you can fully dismiss weird Bay Area people. So that’s why we are seeing the pivot towards the personal attacks on these people.
This stuff? A spanish speaking critic of AI safety says, "Hey guys, it's really not so dangerous." What did you mean by "this stuff" precisely? Is it like "you people" perhaps?
Absolutely. Addiction is an illness. It just happens to be one that is more publicly visible than most.
I empathise with everyone I'm able, and recently struck up a friendship a Nigerian scammer as a result. After jerking him around for awhile, I broke him out of his script. Rather than blocking me or cursing me, I found a moment of sincerity. I told him that I understand the desperation of his situation, but that the people he was most likely to scam successfully would be others in similarly difficult situations.
Suddenly, I was talking to a real person. A young man in an impossibly poor country with few other options...facing the loss of his already meager accommodations and the prospect of begging on the streets. A young man whose heart still hurt him when he attempted to exploit others. One who'd not yet been so jaded by years of such activity. I have cautiously continued talking to him and learning about his situation. He's respected every boundary I've placed to protect myself, and not pushed me to give him anything, even preemptively rejecting what he interpreted as an attempt to send him gift cards at one point, when he misread the intent of my questions about the mechanics of the scam he had tried on me.
I learned he is quite skilled at designing and producing beautiful traditional garments, but has been unable to find a work that would allow him to take care of himself.
I'm now seeing if people I know might be interested in ordering his clothing, and if he successfully delivers the garments they commission from him, I'll build and host a website for him to begin selling online and work to help him find other customers. It may not only lift him out of abject poverty, but give him to opportunity to elevate others and save them from the same spiral of despair in which he found himself...which would mean fewer grandmothers and vulnerable people being scammed in the future.
Granted, he could be running a long con and masterfully manipulating me at every turn...but even if so, I won't regret choosing empathy over judgement and anger, if only for the chance that it might make the world a slightly better place...even if just for one young man in a small city on the other side of the world.
I'm far less interested in how good a big expensive model is on hardware 99% of people can't afford and would rather see what runs best on a chromebook or mobile phone with 8GB of RAM.
The card in question here had an initial MSRP of $1600. It's been bumped up by the market, probably because it turned out it's nice for things like this, but it's hardly in the 99% can't afford domain, especially if you're using it to replace a never-ending rent at which point it will pay for itself very rapidly, especially for heavy LLM users.
In any case, we've gone from requiring supercomputers, to requiring very high end computers, to requiring $1600 video cards. It's tracking the exact same path that image rendering systems took (which if you haven't been keeping up there, now run excellently on pretty much any plain old computer), and we'll probably be there within a couple of years if not much sooner.
You misread your article, which really shouldn't pass a sniff test. 77% could cover the $400, with 14% of that being from people selling/credit. It also mentioned that 70% have enough saved away to cover all expenses for at least 3 months, including 15% that could do so by selling/credit. Also the US is far from the richest country in the world in terms of the population. We're 28th in terms of median wealth. [1] And that's nominal - measure it PPP and we'd be much lower. For the times they are a-changin'.
It was available at this MSRP 3 years ago, direct from Nvidia. It now goes for 3-4k USD, as you point out. MSRP stopped being a useful value for graphics card around that time. You realize the price hike, but still mentioned that 1.6k figure after.
No normal person is spending 3-4k on a GPU from 3 years ago. The availability is also of questionable provenance.
The MSRP is a good proxy for the 'level' of a card. It's not like the 4090 was a freak outlier. Cards of a comparable price offer comparable performance. So the 'level' for running a frontier level model at high performance is now at $1600 and continuing to trend sharply downwards.
Another nuance is that the computer hardware market is currently extremely inefficient in a way I don't understand. You can pick these cards up locally at places throughout Asia for around $2k new. That's retail single unit prices. No idea what's stopping somebody from closing the gap and making a ton of money - perhaps tariffs and data centers purchasing in a price insensitive fashion. Whatever the exact reason may be, what people pay for hardware is increasingly just radically different depending on where you buy it at.
That arbitrage opportunity is remarkable. Maybe duty/taxes, as you say?
I'm the guy who (Who plays games and runs molecular dynamics simulations and other CUDA stuff) said 3 years ago "$1600 for a graphics card? That is excessive. I'll upgrade in a few years when ready" And bought a 4080 for $1200 from Nvidia instead of the 4090. Oops! Now there is no reasonable upgrade path.
The reason models running on low vram are not talked about enough is because they are just not worth it. Qwen3.8 27b changed that, but even 24gb vram is too low for it. Running better model faster at 12gb vram is where its now at, and thats why you see people talkin about it
I might try running the expert pruned Coder model but yes, that PrismML Bonsai 2 Ternary 27B model is from the Qwen 3.8 27B model, which has better intelligence density (Artificial Analysis says), without the MoE disk use or architecture complexity (if you care about that)! There are also DFlash 2 models for it too (though in my experience this only measured faster for parallel requests, but I have a 3090). I am curious about the phone acceleration for Bonsai 2!
Because that's not possible (to have a GPT 5.6 Sol level model). People won't believe this and will keep dreaming, but intelligence is not free and 8GiB (shared with OS and other processes) is too small to be useful. Whether it is possible for 48GiB or 64GiB (meaning useful for model would be ~16GiB to 24GiB) with external fast storage (SSD), OTOH, is a question mark.
I think the problem is not what I don't like about WebComponents individually. The problem is that to use WebComponents, you have to use JavaScript anyway and if it does not do exactly what you wanted, then you're right there, in JS, ready to DIY the problem away.
If WebComponents were purely declarative, I would try harder to use them.
I don't understand why browser makers don't just steal the API from popular solutions like React honestly. That's "fair use" since Oracle v Google afterall.
reply