Hacker Newsnew | past | comments | ask | show | jobs | submit | darksaints's commentslogin

Forgot that the cybertruck was a sales flop and quality joke, and that the Tesla Semi is now the elephant in the room.

The Tesla Semi was groundbreaking when they revealed it nearly 10 years ago. But now there are dozens of electric truck models, and they get delivered in substantial quantities for over a year now. At least in Europe.

But... the Roadster!

This is absolutely amazing.

For those of us programming nerds that want to play with aerodynamics, I can't recommend AeroSandbox enough. While the code is pretty obviously written for people who know their way around aerodynamics and not so much around programming, it is remarkably powerful. You can do all sorts of aerodynamic simulations and is coupled with optimization libraries that allow you to do incredible aerodynamic optimizations. It comes included with some pretty powerful open weight neural network models that can do very accurate estimates of aerodynamic characteristics of airfoils in a fraction of the time that top tier heuristic solvers (like xfoil) can do (which are already several orders of magnitude faster than CFD solvers).

https://github.com/peterdsharpe/AeroSandbox


Thank you for the kind words! :)

No, felon illegals are supposed to go to prison and serve their sentence before deportation. You know, because they committed felonies.

Crossing the border illegally is a misdemeanor and overstaying your visa is a civil violation and not even a crime. Yet somehow those are the only ones he's targeting. Those, and actual lawful immigrants that say things that he doesn't like.


They are deporting people after release from prison who weren’t deported like they were supposed to be. How do you not know this?

wow.dhs.gov

I understand where you're coming from, and I'll play devil's advocate to the devil's advocate: If generative AI is generating convincingly photorealistic CSAM, what the fuck are they training the models on? And if those algorithms are modifying images of actual children, wouldn't you consider those victims?

I strongly sympathize with the idea that crimes should by definition have identifiable victims. But sometimes the devil doesn't really need an advocate.


Considering that every image generation model out there tries to censor your prompts/outputs despite trying their best not to train on CSAM... you don't need to train on CSAM for the model to be capable of generating CSAM.

Not saying the models don't get trained on CSAM. But I don't think it's a foregone conclusion that AI models capable of generating CSAM necessarily victimize anyone.

It would be nice if someone could research this, but the current climate makes it impossible.


> If generative AI is generating convincingly photorealistic CSAM, what the fuck are they training the models on?

CSAM of course: https://www.theverge.com/2023/12/20/24009418/generative-ai-i...

When you indiscriminately scrape literally billions of images, and excuse yourself from vigorously reviewing them because it would be too hard/expensive, horrible and illegal stuff is bound to end up in there.


That's probably incidental, horrible as it is. Models don't need training data of everything imaginable, just enough things in combination, and there's enough imagery of children's bodies (including non-sexual nudity) and porn to generate a combination of the two, same as it can make a hybrid giraffe-shark-clown on a tricycle despite never seeing that in the training data before.

The biggest issue here is not that models can generate this imagery, but that Musk's Twitter is enabling it at scale with no guardrails, including spamming them on other people's photos.


Yep, when my kid was taking selfies with my phone and playing with Google Photos, I appreciated that Google didn't let any Gemini AI manipulation of any kind occur, even if whatever they were trying to do was harmless. Seemed very strict when it detected a child. Grok should probably do that.


Do you need photos of humans to create photorealistic inappropriate material? Could it be derived from 3D art / cartoons?


>If generative AI is generating convincingly photorealistic CSAM, what the fuck are they training the models on?

Pretty sure these models can generate images that do not exist on their training data. If I generate a picture of a surfing dachshund, did it have to train on canine surfers?


warning: incoming generational stereotypes

The original phrase "The customer is always right" had an important caveat: "... in matters of taste". Somehow boomers managed to forget the caveat and created a culture of treating customer service workers like personal slaves and demanding to be treated like royalty. I don't know that Millenials think the customer is always right, but I do see that the Zs think anybody can be wrong, especially customers, and I love that about them.


> The original phrase "The customer is always right" had an important caveat: "... in matters of taste".

This is not true. This is fairly recent Internet revisionism with no historical basis, an attempt to "well ackshually" the phrase into something else to make a point.

"The customer is always right" has always meant, "It's better to appease occasional assholes than to risk ever disappointing a customer with a legitimate grievance." Of course, TCIAR is not a natural imperiative. It's an unproven philosophical proposition that may or may not be appropriate at any given time and in any given industry.

To the extent that there is an "original" formulation, it might be Marshal Field's "Give the lady what she wants," which is (probably intentionally) vague on exactly how it should be applied. (And remember that "lady" in Field's time had some classist connotations that are less present today.")


honestly at this point I'm actually surprised that there aren't specialized linux distributions for hosting postgres. There's so many kernel-level and file-system level optimizations that can be done that significantly impact performance, and the ability to pare down all of the unneeded stuff in most distributions would make for a pretty compact and highly optimized image.


Honestly, I don't have much faith in Linux anymore, and it has everything to do with the explosion of the kernel's codebase due to the explosion of cheaper devices running linux and the (admittedly difficult) management issues surrounding the kernel. I feel like from a security perspective, macos is a better choice and that pains me as a long time linux user.

Can we please move on to microkernels already? I'm fine with a tiny performance hit, I just don't want to get rooted because I plugged in the wrong USB stick.


You can use microkernels whenever you want. Just be aware that they typically have the same issues with zombie/cruft code and aren't necessarily more secure for every application.


I think the point is that even drivers could be non-trusted and live outside of the kernel and just provide the exact service required with minimal access.

That said, why do we still need drivers in 2025? Most regular printers should be dumb, U-MASS should be dumb, webcams should be dumb, monitors are dumb, etc... very few devices coming really needs custom drivers anymore (even with many customizations we could provide class specific descriptors that drivers could adhere to).


If you don't want to go macOS route and want to leave Linux world, your destination would be FreeBSD or OpenBSD.

On the other hand, if you're not running Wine, you can't get autorun virii from USB drives, plus the Windows virii just lives there and can't do anything.


What about plan9? ;)


Plan9 is like ocean yacht racing. If you have to ask about the "cost" you aren't the target market.

Plan9 is like writing. You either do it, or talk about doing it. I'm talking not doing btw. I tried, but I got stuck on trivial things and the barrier to asking for help over 2+2= is high. (No offence intended. The 9 heads aren't interested in running a kindergarden)


I want a 4TB SSD.

To do that on a MacBook I'm spending a minimum of 3200$.

If you have unlimited money ( or can expense it) a 3200$ to 4k MacBook is going to be the best experience money can buy.

If you have limited funds, a 200$ used computer can get the job done with the right distro.


I've also felt very similar, and adopted Rust for those reasons. Even though I'd still love to have a garbage collected, AOT compiled ML-variant, using OCaml still feels like a step backwards now. OCaml's module system and generics feel clumsy and unintuitive now that I've used Rust's Trait system.


I have one of these cars, the last model year before they switched to carplay. The bluetooth implementation is dogshit...randomly speeding up songs, audio skipping, cutting out, etc. Occasionally forgets connections and has to be set up from scratch. Can't handle music and navigation at the same time...will turn down volume on music to handle navigation instructions, only to forget to turn the music back up for 5 minutes, by which time you've already turned up the volume to compensate and get blasted by the unexpected volume increase. And from what I can tell, my car does it better than most other manufacturers from the same era.

I think the only reason why the car makers switched to carplay/android auto is because they knew they sucked at infotainment systems and software and nobody wanted their bullshit, so they just outsourced it so they wouldn't have to try anymore.


That situation is ridiculous, my 15 year old diesel has an extremely bare bones touchscreen radio/info dashboard. The Bluetooth, old as it is, works fine.


> The bluetooth implementation is dogshit...randomly speeding up songs, audio skipping, cutting out, etc. Occasionally forgets connections and has to be set up from scratch.

We need to go back to the days of headphone jacks and analog AUX ports. Seriously. Something simple, reliable, and extremely well-understood.

Cars should be a lot more modular, but the market is consumer-hostile in general and it won't happen without regulation. There should probably be some standardized car-infotainment interface, and consumers be required to buy the infotainment system aftermarket to get more competition and quality in that area.


Are there any promising core designs yet? Multi-core designs? Any promising extensions being standardized?

I really want to believe, but I don't think we'll see anything like an M5 chip anytime soon simply because there's so little investment from the bigger players.


Yeah Rivos apparently taped out a high performance server class core (probably only a test chip I'd guess) before Meta bought them.

There are plenty of multi core designs (that's easy) but they aren't very fast.

In terms of open source XiangShan is the most advanced as far as I know. It's fairly high performance out-of-order.

I don't think there's anything M5-level and probably won't be for a while (it took ARM decades so it's not a failing). I doubt we'll see any serious RISC-V laptops because there probably isn't demand (maybe Chromebooks though?). More likely to see phones and servers because Android is supporting RISC-V, and servers run Linux.

In terms of extensions I think it's pretty much all there. Probably it needs some kind of extension to make x86 emulation fast, like Apple did. The biggest extension I know of that isn't ratified is the P packed SIMD one but I don't know if there's much demand for that outside of DSPs.


Tenstorrent has announced Ascalon development boards TBA 2026Q2.

That's not gonna beat the M5, but it should be similar or better relative to M1, and a huge performance jump for RISC-V.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: