HN user
I loathe their bug-ridden UI.
So weird that the same exact people telling you that programming careers are now obsolete are the same group who haven't been able to fix screen flickering bugs for like a year...
searching for .gov reveals 0 matches... doubt
very loudly stating the foreign labs have been distilling their models
Help! Someone else is blatantly ripping off my plagiarism machine!
Fail cleanly.
This is the same exact industry that gives you paid usage limits as a unit-less percentage bar then gaslights customers every time the algorithm running that percentage bar changes or they lobotomize an existing model with increased quantization to squeeze a few more dollars out of existing hardware.
"Failing cleanly" might make their moated hype-machine look bad pre-IPO, so they certainly aren't going to do that voluntarily.
Yes, but they are suuuuper safe. /s
So far I have mixed impressions, but they do indeed seem noticeably weaker than comparably-sized Qwen3 / GLM4.5 models. Part of the reason may be that the oai models do appear to be much more lobotomized than their Chinese counterparts (which are surprisingly uncensored). There's research showing that "aligning" a model makes it dumber.
That is fine as long as the input / output is always in UTC... but at the end of the day you often want to communicate that timepoint to a human user (e.g. an appointment time, the time at which some event happened, etc.), which is when our stupid monkey brains expect the ascii string you are showing us to actually make sense in our specific locale (including all of the warts each of those particular timezones have, including leap second, DST, etc.)
That’s a good target for terrorism.
Exactly! Nobody would be laughing if Al-Qaeda drove a giant Dasani truck into TSA headquarters, would they?
Iirc. you just need the openMP libraries from a compatible version of clang.
Agree 1000%. I'd take less pay to not have to deal with some middle-management dimwit who read this article trying to actively "coach" me.
to do what boils down to a few API calls
LOL @ anyone who believes that global financial processing is primarily a technical problem vs. the regulatory / bureacratic dystopia it actually is.
Well Google's garbage-tier product cohesiveness makes a LOT more sense now.
IMHO, the overhead of perpetually babysitting compiler diagnostics or performance metrics to ensure your latest update didn't confound the auto-vectorizer is never a net positive over just using something like xsimd, Google highway, etc.
Agree 100%... Getting caught using a phone while driving, should be punishable by a suspended 5-year prison sentence contingent upon completion of a 1 year smartphone ban. Get caught using anything other than a flip phone at any point during the next year and you have to serve the prison term.
thief breaking into your house, stealing all. . .
This is where your analogy is flawed. You are pre-supposing the "defendant" is indeed the thief that stole your property. Whereas that is entirely a legal determination which is the outcome of a trial AND at the heart of this discovery request. More aptly if you thought steve stole your red Ryder bb gun, and Steve was indeed found to be in possession of a red Ryder bb gun, it would still be the prosecution's burden to prove that Steve stole it from you (instead of purchased it from a store).
Similarly here, if NYTimes is claiming that openai's gpt4 illegally reproduces "to be or not to be. . ." (Or whatevs) from issue #8628 page 76, it's still their burden to prove that is actually a thing that is both copyrightable and that they own the copyright to vs. openai just reproducing Hamlet instead of a nytime's reporter's particular review of a production of hamlet in that issue. Etc. etc.
More germanely, if you point an llm at a pile of source documents and ask it to write a newspaper article, it'll happily do so in 2024. Understanding if/how this is fundamentally different from what a reporter does when synthesizing that same article goes to the very heart of this case (i.e. which transformative works are indeed copyrightable)
Speaking of sugar coatings, have you tried Candy Crush Saga 365 yet? I've already pinned it in your start menu favorites for convenience.
That’s how admissions used to be run
ish... TBF, now you have to do an additional secret indirection dance to get around the Supreme Court's ruling on affirmative action.
No device support...
I started playing around with porting some CUDA code to ROCm/HIP on a Ryzen laptop APU I had. While an "unsupported" configuration (which was understood), it all worked until AMD suddenly and explicitly blocked the ability to run on APU's. Currently the only way to get back to work on that project on that particular computer would be to run a closed-source patched driver from some rando on the internet. Needless to say, I lost interest.
Last I checked, there were only 7 consumer SKU's that could run AMD's current compute stack, the oldest being 1 generation old. Even among the enterprise hardware they only support ~2 generations back. So you can't even grab some old cheap recycled gear on e-bay to hack on their ecosystem.
Meanwhile, I can pull anything with an NVIDIA logo on it from a junkyard it'll happily run CUDA code that I wrote for the 8800GTX 15+ years ago.
Simply register and enable the repositories that you need.
Counter offer: nah... how about I just target+deploy on any of the numerous competitors that don't make me jump through ANY licensing hoops whatsoever from here on out.
Almost every major open-source project is actively contributed to by Red Hat, or has been at one point.
That's nice... but Redhat's entire business is also built on the open source software contributed by tens of thousands of others, much of it licensed to them under copyleft terms like the GPL. Yes, RedHat absolutely did contribute immensely to that ecosystem, but once they (or corporate-daddy IBM) decided to take a big fat stinky dump in the collective sandbox and stopped sharing their toys, the rest of us are kinda allowed to be pissed at them, no?
"We will give you the SRPM because we legally have to, but if you actually exercise any of the rights afforded to you by the GPL that software was licensed to us under we will immediately terminate you as a customer" *may* (a court will ultimately decide) fit into some legal loophole that exists in the void between contract vs. copyright law, but it certainly does violate the spirit of term#6 of the GPL, i.e. "You may not impose any further restrictions on the recipients' exercise of the rights granted herein".
RedHat pulled what is commonly known as a "dick-move". Ef them and the horse they rode in on.
Meh... there are plenty of (better) fish in the "do your own thing" sea. The thing that actually made CentOS, followed by Rocky+Alma compelling to anyone was the 1:1 bug compatibility with RHEL. Not sure what use-case exists for Alma once they are nothing beyond a mere knock-off of CentOS stream. Like why would you ever pick *that* over literally anything else?
which states you may not impose any further restrictions on the recipient's exercise of the rights granted herein.
So "technically" that's not what they are doing. Redhat will happily give any of their customers the source to any GPL binary that customer received from Redhat. This does satisfy the GPL license requirements. Redhat will then immediately terminate their contractual relationship with that customer (i.e. we owe you no more binaries and therefore no more sources AFTER today) if that customer does indeed re-distribute that SRPM.
In other words they are not imposing any restrictions on this specific GPL binary/source pair. They are just ending their contract with you and cutting you off from all future redhat binaries AND sources.
I believe this is a bullshit end-run around the GPL, but whether this intersection of contract law and copyright law actually represents a "restriction" as defined in the GPL is going to 100% be up to a court to decide.
it doesn’t just throw up its hands and disengage - it’s going to attempt to stop or maneuver out of the way of danger
Lol. That is EXACTLY what every modern auto driving system I've used does. If it gets confused it beeps and then immediately bitches out.