I'm not sure that's entirely true. For most things, checking if a solution is correct is much easier than implementing it (page looks wrong, can't login etc...)
I think some of it might be genuine. For people that don't code (like management), going from 0 to being able to create a landing page that looks like it came from a big corporation is a miracle.
They are not able to comprehend that for anything more complicated than that, the code might compile, but the logical errors and failure to implement the specs start piling up.
This paper creates a new benchmark comprised of real remote work tasks sourced from the remote working website Upwork. The best commercial LLMs like Opus, GPT, Gemini, and Grok were tested.
Models released a few days ago, Opus 4.6 and GPT 5.3, haven't been tested yet, but given the performance on other micro-benchmarks, they will probably not be much different on this benchmark.
One of the tasks was "Build an interactive dashboard for exploring data from the World Happiness Report." -- I can't imagine how Opus4.5 could've failed that.
Waiting until the moment they get good enough is not a smart thing to do either. If you are a farmer and know it is going to snow, at some point in the next 5 months, you make plans NOW, you don't wait until the temperatures drop and you see the snow falling. Right now, people are waiting for the snowfall before moving their proverbial chickens indoors
Top AI researchers like Yann LeCunn have said that LLMs are a dead end.
It seems to me that LLM performance is plateuing and not improving exponentially anymore. This recent hubbub about rewriting a worse GCC for $20,000 is another example of overhype and regurgitating training data.
You don't know for sure if it is going to "snow" (AI reaches general intelligence) Snow happens frequently, AI reaching general intelligence has never happened. If it ever happens, 99% of jobs are gone and there is really nothing you can do to prepare for this other than maybe buy guns and ammo, and even that might not do anything to robotic soldiers.
People were worried about AI taking their jobs 60 years ago when perceptrons came out, and anyone who avoided a tech career because of that back then would have lost out majorly.
There is no reason why an AI model capable of pushing a significant chunk of devs into lower paid and highly competitive dev jobs as a result of automation needs to be a general artificial intelligence. There is a lack of nuance that comes with thinking that either AI is dumb or it has human level general intelligence. As much as devs hate to admit it, you don't need that much of what we understand as general intelligence to write software. Only a portion of your intelligence is needed and arguably not all of it at the same time.
While general purpose models might be plateauing soon (arguably they have for a while). Highly specialised models (especially for programming) haven't necessarily plateaud yet. And anyway, existing functionality seem like a good foundation to build upon systems that remove the need of hiring as many devs. It's not the "being out of a job" that should worry you. Open up your binary thinking and consider that facing a 08 job market for the rest of your career is not the same permanent unemployment but it is not a market you would like to have.
You don't need to be a genius or rocket scientist to write code, but llm don't even reach the bar for anything but the most simple things. Take a look at the video I posted earlier for an example.
And specialised models for programming HAVE plateaued.
> Can you imagine not being fired when you can only do 2.5% of all tasks?
You are not competing against LLMs though. You are competing against people (who in a pre-LLM world wouldn't be in tech) using LLMs tools to beat you in terms of value. In the new world, you either are a top 1% dev or you beat everyone in race to the bottom pricewise. The middle will become vanishingly small. Think of manufacturing in developed countries.
This technique showed that there are ways during training to optimize weights to neatly quantize while remaining performant. This isn't a post training quantization like int4.
For Kimi quantization is part of the training also. Specifically they say they use QAT, quantization aware training.
That doesn't mean training with all integer math, but certain tricks are used to specifically plan for the end weight size. I.e. fake quantization nodes are inserted to simulate int4.
Iirc the paper was solid, but it still hasn’t been adopted/proven out at large scale. Harder to adapt hardware and code kernels to something like this compared to int4.
Has anyone noticed a lot of polymarket posts on their X (formerly known as twitter) feed claiming to be making a fortune? It makes me feel like its some kind of coordinated guerilla marketing scheme.
Perhaps that's because there's an enormous difference between fine art and computer programs.
Also, there's quite a lot of pushback against AI-generated code, but also because unlike music, normal people have no interest in and aren't aware of the code.
They are obviously different things, but aren't the people who spent thousands of hours honing their coding and releasing their code spending just as much time and effort if not more than the people who made non-ai images and music?
There certainly are. If you by chance imply equivalence based on time and effort spent, neither is the thing that differentiates arts, crafts and other activities.
I won't merge anything AI generated in any of my FOSS projects, unless I'm successfully deceived.
In the first place, I do not regard a copyright notice and license on AI generated code to be valid in my eyes, so on those grounds alone, I cannot use it any more than I could merge a piece of proprietary, leaked source code.
The copyright office agreed with you about the non-copyrightability of AI generated media so in that sense you can safely ignore copyright claims on anything AI-generated.
Music is art, code is engineering. "Hackers and painters"[1] was always wishful fluff, unfortunately.
When it comes to code, I don't think anyone cares how the sausage is made, and only very rarely do people care by whom. The only question is "does it work well?"
Art is totally different. Provenance is much more important - sometimes essential. David is a beautiful work, but you could 3d print or cast a replica of "David". No one would pretend that the copy is the same as the original though - even if they're indistinguishable to the untrained eye - because one was painstakingly hand sculpted and the others were cheaply produced. This sense of provenance is the property that NFTs were (unsuccessfully) trying to capture.
If someone painstakingly hand sculpted an exact replica of "David", does it make it art, or a forgery? Is hand written code to produce generative art not art?
It's difficult to pin down the line. Ultimately it's up to the individual to define them. "The relationship to art, and this kind of painting, to their work, varied with the person entirely."[1]
> If someone painstakingly hand sculpted an exact replica of "David", does it make it art, or a forgery? Is hand written code to produce generative art not art?
No and no.
If you raise and teach a child and they generate a painting, are you the artist?
Devs are quite used to using others peoples work for free via packages, frameworks and entire operating systems and IDE’s. It’s just part of the culture.
Music has its history in IP, royalties, and most things need to be paid for in the creation of music or art itself.
It’s going to be much easier for devs to accept AI when remixing code is such a huge part of the culture already. The expectation in the arts is entirely different.
This doesn't make sense to me. I mean, the term "remix" literally comes from the music scene.
Artists are constantly getting inspiration from one another, referencing one another, performing together or having their works exhibited together...
While there are some big name artists who are famously protective of the concept of IP, those artists have made headlines exactly because when they litigate they seem so unreasonable compared to the bedroom musicians and pub bands and church choirs and school teachers and wedding DJs and millions of other artists and performers whose way of participating in "the culture" is much less tied to ownership.
Most code people interact with are creations shat out by soulless corporations, why would they care? Being honest here, the vast majority of people have their code experience dictated by less than a handful of companies; at their jobs they are told to use these tools or get file for welfare. The animosity has been baked into the industry for quite a while, it's only very very recently that the masses have been able to interact with open source code and even that is getting torn down by big tech.
Compare this to music where you are free to choose and listen to whatever you want, or stare at art that moves you. IF you don
At work most people are force to deal with code like SalesForce or MSFT garbage, not the same experience at all.
Why would people care about code coming from an industry that has been bleeding them dry and making their society worse for nearly 20+ years?
Every thread on HN that touches on the topic has countless people talking about how LLM generated code is always bad, buggy and people that utilize them are inexperienced juniors that don't understand anything.
And they're not completely wrong. If you don't know what you're doing, you'll absolutely create dumster fires instead of software
Sure, I am one of the people who will say that. But where are the people calling for it to be banned? Where are the stores and websites that are banning AI generated software?
I feel like part of the difference is how art vs code is viewed. You could make the argument code is art, though most don't have that stance. Visual art and music tend to be made by a few people, there is ego involved, you care who the artist is. Code tends to be made by shops and consumers don't know who the coders are. Programmers are already faceless.
I think it's also about money. Places code and code samples are stored tend to be large companies that are in tech and on the AI hype wagon. Bandcamp is not one of those places.
There's one popular platform that requires disclosing whether and how AI was used (Steam), and if you search anything about it, all you can find is like a sea of articles opposing it.
That's not really the same as stores outright banning AI code.
An apt analogy would be like a shared drawing taking merge requests and having to spend 30 minutes looking at every single merge request zoomed in to see if there was a microscopic phallus embedded somewhere.
It is completely fair for an open source project to have their own standards, and you are also free to fork it so you can accept as many AI PRs as you want.
None of these options are available for someone that wants to sell AI generated music. There are really only 2 marketplaces to sell your own music and if both of them banned AI, then you are effectively locked out of the entire market.
I think a key factor there is that programmers (in the actual sense, rather than so-called “vibe coders”) are more likely on average than (current) artists and musicians to have intimate knowledge of how AI works and what AI can and can't do well — and consequently, the quality of their output is high enough that it's harder to notice the use of AI.
Eventually that'll change, as artists and musicians continue to experiment with AI and come up with novel uses for it, just as digital artists did with tablets and digital painting software, and just as musicians did with keyboards and DAWs.
AI music from suno sounds indistinguishable to non-ai generated music to me.
In terms of how well it works, the quality of AI music is far better than art or code. In art there are noticeble glitches like multiple fingers. For code, it can call non existent functions, not do what it is supposed to do, or have security issues or memory leaks. From what I can tell, there is no such deal breaker for AI music.
The tells in music are there. The most common being: vocals have a subtle constant hiss to them, voices and instruments sound different in the second half than they did in the first, the hiss filter gets more prominent and affects all instruments towards the end of the song, auditory artifacts like volume jumps or random notes/noises near transitions.
More subjective tells: drums are hissy and weak, lyrics are generic or weird like "Went to the grocery store to buy coffee beans for my sadness", weirdly uniform loudness and density from start to finish, drops/climaxes are underwhelming, and (if you've listened to enough of them) a general uncanny feel to them.
I've generated about 70 hours of AI music and have listened to all of the songs at least once, so it's become intuitive for me to pick them out.
> For code, it can call non existent functions, not do what it is supposed to do, or have security issues or memory leaks.
I guess what I'm getting at is that, since programmers are typically more inclined than the average person to understand how AI works, programmers are therefore ahead of the curve when it comes to understanding those pitfalls and structuring their workflows to minimize them — to play to the strengths and weaknesses of LLMs. A “fancy” autocomplete v. a “fancy” linter v. something pretending to be a junior programmer are all going to have very different rates of success.
The issue hindering art and music is that most people using generative AI for art and music are doing so analogously to the “something pretending to be a junior programmer” role instead of the “fancy autocomplete” or “fancy linter” roles. That is: they're typically using AI to generate works end-to-end, whereas (non-vibe-coder) programmers are typically using AI in far narrower scopes, with more direct control over the final output. I think the quality of AI-based art and music will improve as more narrowly-scoped AI-driven workflows catch on among actually-skilled artists and musicians — and the result will be works that are very different from existing works, rather than works that only cheaply imitate some statistical average of existing works.
Lower cost usually means lower quality and is an example of how a long path being leaked can result in traffic flowing away from high quality path to the leaked path.
Not saying that this is the case with Venezuela, just explaining the reality of BGP where path prepends are often ignored.
I've actually tried hetzner on and off with 1 server for the past 2 years and keep running into downtime every few months.
First I used an ex101 with an i9-13900. Within a week it just froze. It could not be reset remotely. Nothing in kern.log. Support offered no solution but a hard reboot. No mention of what might be wrong other than user error.
A few months later, one of the drives just disconnects from raid by itself. It took support 1 hour to respond and they said they found no issue so it must be my fault.
Then I changed to a ryzen based server and it also mysteriously had problems like this. Again the support blamed the user.
It was only after I cancelled the server and several months later that I see this so I know it isn't just me.
Why wouldn't microsoft advertise this though? If they had the ability to take the attack and others might not, then it'll result in more customers for them.
