What makes you think the next generation models won't be explicitly trained to prevent this, or any other pitfall or best practice as the low hanging fruit fall one by one?
A million times, this. Sometimes they luck into the intent, but much more frequently they end up in a ball of mud that just happens to pass the tests.
"8 unit tests? Great, I'll code up 8 branches so all your tests pass!" Of course that neglects the fact that there's now actually 2^8 paths through your code.
if you can steer an LLM to write an application based on what you want, you can steer an LLM to write the tests you want. Some people will be better at getting the LLM to write tests, but it's only going to get easier and easier
This is one of the reasons why I just wrote a testing book (beta reviews giving feedback now). Testing is one of those boring subjects that many programmers ignore. But it just got very relevant. Especially TDD.
No, OP is merely an AI deepthroater that will blindly swallow whatever drivel is put out by AI companies and then "benchmark" it by having it generate a pelican (oh and he got early access to the model), then call whatever he puts out "AI optimism"
The reality of things is, AI still can't handle long running tasks without blowing $500k worth of tokens for an end result that doesn't work, and further work is another $100k worth to get nothing novel.
Where are you pulling these numbers from? I'm genuinely interested. Is it the kind of budget you need to spend in order to have Claude build a Word clone?
Mark Zuckerberg explicitly called out the airpod pairing being closed as unfair in a semi recent interview, maybe he can throw some dollars that way and get it all working nicely in some meta products.
It's not AirPods being closed that's unfair. Apple should be able to sell first party tech that only works with their own products.
What's unfair is Apple locking everyone else out. Not allowing or documenting for third parties to use the same APIs to enable something like automatic device switching in third party bluetooth headphones is the unfair part.
Same goes for the watch. That the Apple Watch only works with iPhone isn't the problem. The problem is no other third party is able to make a smartwatch that competes on an level playing field with the Apple Watch on Apple Devices, because Apple locks them out.
By locking their products "in", they're also locking third parties "out". How on earth would they be able to "lock-in" the Apple Watch to the iPhone while at the same time NOT "locking-out" third parties?
I think they're suggesting that lock in implies apple didn't write the code to help support 3rd parties. Lock out implies they actively wrote code to prevent 3rd parties.
With other companies there may be a difference. But with Apple, for all intents and purposes, it's the same thing. Because they are hostile to third-party integration using undocumented API's or interfaces.
I'm not the person you're asking to but this is my reasoning:
1. If I'm building a gadget for my line of products, I want to be able to test it only with my products. I don't want to spend money to make it work with anybody's else products. There are standards but there are bugs and non compliant products from known and unknown parties, their problems.
2. However I might also want to be able to build gadgets for somebody's else products, so I appreciate if those companies stick to standards and don't go out of their ways to make their products incompatible with gadgets of third parties. BTW, this reminds me about cartridges for inkjet printers.
So I think that it would be fair for Apple to say, "these earpieces are tested to work only on these products of mine: ...; if they happen to work on something else: congratulations! you got lucky."
It won't be fair if they make their products incompatible with every other earpieces and at the same time claim that they are compliant to a standard.
…unless someone has sufficient time and/or money to spend on it, and wants to do so as a point of principle.
If I had large amounts of spare money, I’d love to seed small endeavours that (according to my personal world view) made the world incrementally better.
As has been noted before, what’s the point of having ‘FU money’ if you don’t use it to say ‘FU’ now and again?
Yeah, but saying "maybe super rich people will do random illogical things" isn't really a great argument. For all we know, Mark Zuckerberg wants to spend his "fuck you" money locking the ecosystem down even more, as a "fuck you" to consumers.
You should focus on the p2p part of code and object distribution. While nix is not perfect, people are not going to learn and adopt yet another package manager.
A distributed git object cache is what is really needed at the moment.
I'm actually working on my own OS agnostic package collection + system management software, and I've found https://radicle.xyz great for this. All repos depended on by the official package collection t will be on the radicle network.
The reasoning for the shooting is pretty clear. He told his transgender lover that “I had enough of his [Kirk’s] hatred. Some hate can’t be negotiated out.”
The school shooting that Mr. Kirk lost his life to is not, "left wing violence". Unless you want to submit that most school shootings are "right wing violence" if the shooter hated public education.
> There are multiple statements from his relatives that his family are "MAGA", his parents are Registered Republicans.
To be fair, that doesn't necessarily say anything about his politics. I know plenty of liberals with MAGA parents. I don't think we can draw any conclusions as to his politics at this time.
Exactly, which was the point Kimmel was making. Apparently that suggestion was too much for the current administration, and the official narrative must not be questioned.
> Most leftists despise their parents[sic] politics.
And do you have a source on that? Anecdotally, most "leftists" I know have left leaning parents. But it's up to the person to define if they are or are not "leftist", because it's a rather narrow, small minded world view that has to define things in those terms.
> None of this suggests a rightward leaning of the culprit himself.
Nor does it suggest his leftward leaning. Maybe it suggests why he used violence as a means to enact social change on the world.
Wow. If you haven't met folks who don't fit, or don't consider themselves aligned with the "left/right" spectrum of American politics, you're missing out. The "us vs. them" mentality is juvenile, and it is sad you subscribe to it.
If you cannot comprehend the shades of grey in the world, maybe you need more exposure to it.
It was well known before Kimmel made his comments that the shooter was in a romantic relationship with a trans woman. Having said that, even if he did not know about that relationship it was irresponsible of Kimmel to repeat rumors he could not have known were true that the shooter was maga.
He did not assert they were unknown: “We hit some new lows over the weekend with the MAGA gang trying to characterize this kid who killed Charlie Kirk as anything other than one of them,” Kimmel said in his Sept. 15 monologue.
I guess that leaves the question as to why Kimmel did not say: 'We hit some new lows over the weekend when people of all political stripes were trying to characterize this kid who killed Charlie Kirk as anything other than one of them'. Because that seems like it gives more information to the viewer because that is what actually happened and acknowledging people from both sides were doing the 'bad thing' should help to bring people together instead of driving people apart.
Mr. Kimmel is making a series of jokes about how members of the party in power are reacting, including a clip of the president not seeming to care when asked about Mr. Kirk's death.
Confused politics isn’t all that unusual; look at Caitlyn Jenner for a concrete example. Add in the usual bad blood between well-armed groups and it certainly happens.
I wish everyone would wait a week for actual reliable info to come out. I wish we weren’t getting a bunch of said info from deeply partisan and untrustworthy fuckwits.
According to the latest iteration, his right-wing family said he was left-wing and even neighbors saw him with his roommate.
Freedom of speech is protected. That people are celebrating a man's death, and worse yet, justifying it, is evil but still protected. But what's not protected is the consequences of these actions. I don't want to live, work, etc... next to someone who thinks that it's ok to commit acts of violence against others just because we don't share the same views.
> But what's not protected is the consequences of these actions.
But this is protected in this case.
I can unfriend you on Facebook for saying “I’m not sad he’s dead”. (And to be clear, Kimmel didn’t even go that far.) I can kick you out of my birthday party. I can complain to your employer. They can fire you. (They can fire you for having tattoos, or red hair!)
But the government cannot do these things. That is the entire point of the First Amendment. The FCC can not threaten the license of a broadcaster for protected speech, but we are here anyways.
The entire point of the 1st Amendment is to protect the citizens from being thrown in jail or being prosecuted for speaking against the government.
Where do you see that here? The FCC chairman just said that
"...broadcasters are entirely different than people that use other forms of communication. They have a license granted by is at the FCC, and that comes with it an obligation to operate in the public interest".
To be fair, the new revanchist right calls actual conservatives "left wing". They call libertarians "left wing". They call the shared American values of the past fifty years "left wing". They call straightforward consensus reality "left wing". They basically call anyone who doesn't subscribe to the extended reactionary cinematic universe "left wing". So the only data point there is that his parents are suffering social media psychosis.
Also, non-normative sexual behavior is more indicative of being a Republican ("I have a wide stance!", etc, etc, etc). Democrats just espouse not beating yourself up over it, whereas Republicans seemingly yearn for the closet.
> next to someone who thinks that it's ok to commit acts of violence against others just because we don't share the same views.
But that still only includes a subset of views?
I mean what you are saying is right. But these people were perfectly fine with ignoring or sometimes outright endorsing political violence until one of their own was the target. That does not seem extremely hypocritical?
My bias in these cases is that the simplest answer, same as any mass shooting, is that the killers motivations are a manifestation of mental illness and nothing more. Not always true but typically so; wasnt the trump would be assassin not left for instance? When i was told that i wasnt surprised, not because i think it was more likely of someone on the right, but because i think its mostly random. Eg we have a gun culture, a toxic culture, and a lack of mental health institutions culture. That will only ever produce (among other things) a consistent stream of random acts of violence.
In this particular case i am a little more curious than usual to find out if that holds up here if only because the narrative was so immediately anti left attacks.
Just look at the guy who shot Trump's ear. He had no discernable motive or explicit political leaning at all. And had supposedly been tracking both Trump and Biden. He just did it for attention.
Are you saying there are no conservatives who are attracted to those who identify as trans? Not too long ago you could say the same thing about being conservative and being attracted to the same sex, yet that isn't something be bat an eye at anymore.
Being trans (or being attracted to a trans person) is one of many aspects of that person. Other political positions may outweigh it, and the taboo nature of it may be an appeal (see also: stepsibling roleplay porn).
Kind of surprised by this take - I use openpilot often and also use claude code.
I kind of consider them the same thing. Openpilot can drive really well on highways for hours on end when nothing interesting is happening. Claude code can do straight forward refactors, write boilerplate, do scaffolding, do automated git bisects with no input from me.
Neither one is a substitute for the 'driver'. Claude code is like the level 2 self driving of programming.
Author here - thanks for linking - Unfortunately I didn't have time to continue on this for a long while. I am certainly happy if people get inspiration from any ideas and continue on with it.
I dare say I'm more familiar with the capabilities of the leading models than certain big tech CEOs are, at least judging by their publicly communicated opinions.
I use the 4o very often in my work and it mostly sucks. Sometimes it’s very good, sometimes it has nice knowledge that was faster to find from it than a search engine. Mostly it spouts out unhelpful noise (for my problems).
I’m sure if you need to make a to-do list in react it’s like magic (until the app gets complicated). In real world use, not so much.
(Also I have often code reviewed PRs from people who are heavy users and surprise surprise - their output is trash and very prone to bugs or being out of spec.)
I also think 4o sucks, but have you tried DeepSeek R1 (free on their website)? I thought it night and day between 4o and o3-mini on the following topics:
- reverse engineering: when fed assembly (or decomp or mock impl), it's been consistently been able to figure out what the function actually does/why it's there from a high-level perspective. Whereas ChatGPT merely states the obvious
- very technical C++ questions: DSR1 gives much more detailed answers, with bullet points and examples. Much better writing style. Slightly prone to hallucinations, but not that much
- any controversial topic: ChatGPT models are trained to avoid these because of its "safety" training
ChatGPT is a bit better (and faster) at writing simple code and doing some math faster, but that's it.
(obviously, common sense about what to share and not to share with these chatbots still apply, etc.)
You can access DeepSeek R1? For me, both chat and API have been down for over a week now (it shut down minutes after I topped up my account and generated an API key - I never got to use it :/).
There's lots of fiddling with these models. I found Claude 3.5 Sonnet to be superior to both GPT-4o and o1-preview in around 99% of the things I do; I only started comparing it against o3-mini, and right now it's a mixed bag. Then again, I tend to develop and refine specific prompts for Sonnet, which I haven't for o1-preview and o3-mini, so that could be a factor. Etc.
Yes, well, I live in the EU and thus can avoid US work hours and Chinese peak hours. I think availability has been a bit better since they disabled websearch (also I noticed DSR1 half a week before it made the mainstream news).
I live in the EU too. For me, the status page[0] shows a continuous API outage for the past 8 days, that is still ongoing. Since it started, my API requests bounce back with an error, which changes seemingly at random between "unauthorized" and "insufficient balance". Neither of those reasons are valid, since I'm using a valid API key I made after creating an account, which I topped up with $20 (and have an invoice from them to prove it). I must have had a mightily bad luck that the service went down soon after I generated the API key - I'm guessing my user/key is currently stuck in the middle of some migration, or possibly wasn't captured in a backup and got subsequently wiped. For now, I'll just patiently wait for them to fix their service.
AFAIK it's hosted on Chutes for free too (though limited to between 2k and 10k output tokens). Azure as well, though it might be ratelimited there (or at least it is through openrouter)
reply