It's training on every single codebase in the world instead now You prompt it to fix the mistakes it made and that goes into the new training data
Pro is better, except for agentic coding work
The nazis in this case being the people that allow a wider range of political opinions on their platform than the people you're defending as "just a little mean" There's a very narrow political line you have to maintain…
Twitter was the best website by far, very high quality posts, per day I would only scroll through 3 or 4 tweets because each would be a rabbit hole (as an example, and he's long left the platform, crypto cuttlefish)
When I tried to get into it, all the biggest instances had policies you had to sign off to to make an account where you basically had to agree with them ideologically on everything, I made a serious effort to find a…
I don't think Nietzsche would complain, although doing it for a living person would be very funny (for example when designing a promotional poster)
Google definitely had the best model at points if you're not only concerned with coding, but then they didn't release any pro models for 8 months
It's appealing to me because it's not just one benchmark but a summary of a collection of benchmarks. Also the current rankings have Opus on top by far and everyone's experience is that Opus is best by far so it's not…
It's not the fault of the users when there's no viable alternative Bluesky is founded by the same founders of Twitter and is not open and has ridiculous policies, Mastodon is the same, even less popular and even more…
As I understand they're being pressured into it by governments like the Brazilian government to prevent a widespread form of scam there, but of course they're happy to comply because it means they have more control as…
So they're a piece of shit because you've decided that surrendering necessarily means losing physical possession I agree with the wording, having to give your ID to Google when we'd all much rather that they don't have…
It's an outdated model that they're increasing the price of so you switch to the newer one essentially, this whole post is clickbait since it doesn't mention that
Saw a post recently about how AI agents were used to check if papers had flaws in their conclusions and it found 15% did, that kind of auditing is already a massive massive advancement for science
I and many people believe that google intentionally chooses to keep harmful ads (deceptive, scamming, or leading to malware) this is because it's often very blatant and often survives being reported
Ah I see, it was a self defensive crime against humanity and act of terror, that's good to know, say this to the children whose arms burnt off
> as for trust, avoid it We all end up running closed source code somewhere If you do you'll be much better off avoiding Israeli tech As a rule if you boycott everything Israeli it'll be smarter, more ethical, and safer
You'll have to read the license agreement for openAI or Google on their tts Or look into open source tts models with liberal commercial licenses Generally I'd stay away from the site/project/repo in the OP and just do…
It probably isn't a good tutor, I'm using frontier SOTA models and whenever I ask them to explain anything they don't consider my question and don't display any understanding but just spam out unrelated simplistic…
How does this compare with audiomass?
What do you mean
I gave it a trivial HTML job and it messed it up in several ways including being lazy and lying about results It's probably very hit or miss like everything with LLMs but I was really surprised it performed that badly
> And before anyone starts thinking about whatabouting, Yandex are based in a country where journalists are openly and publicly assasinated to intimadate Thankfully Gary Webb is here with us today to laugh at this
What's worrying for them is nothing they're doing is impossible to be recreated, if you bake it in and you don't want to pay rent you can fine tune your own model
And the only end result would be that the Chinese trained on their data just like OAI and Anthropic trained on our data so who cares
They downloaded pdfs (shock and horror) to create and publish the first open weights LLM Your ethics shouldn't be dictated by US copyright law
It's training on every single codebase in the world instead now You prompt it to fix the mistakes it made and that goes into the new training data
Pro is better, except for agentic coding work
The nazis in this case being the people that allow a wider range of political opinions on their platform than the people you're defending as "just a little mean" There's a very narrow political line you have to maintain…
Twitter was the best website by far, very high quality posts, per day I would only scroll through 3 or 4 tweets because each would be a rabbit hole (as an example, and he's long left the platform, crypto cuttlefish)
When I tried to get into it, all the biggest instances had policies you had to sign off to to make an account where you basically had to agree with them ideologically on everything, I made a serious effort to find a…
I don't think Nietzsche would complain, although doing it for a living person would be very funny (for example when designing a promotional poster)
Google definitely had the best model at points if you're not only concerned with coding, but then they didn't release any pro models for 8 months
It's appealing to me because it's not just one benchmark but a summary of a collection of benchmarks. Also the current rankings have Opus on top by far and everyone's experience is that Opus is best by far so it's not…
It's not the fault of the users when there's no viable alternative Bluesky is founded by the same founders of Twitter and is not open and has ridiculous policies, Mastodon is the same, even less popular and even more…
As I understand they're being pressured into it by governments like the Brazilian government to prevent a widespread form of scam there, but of course they're happy to comply because it means they have more control as…
So they're a piece of shit because you've decided that surrendering necessarily means losing physical possession I agree with the wording, having to give your ID to Google when we'd all much rather that they don't have…
It's an outdated model that they're increasing the price of so you switch to the newer one essentially, this whole post is clickbait since it doesn't mention that
Saw a post recently about how AI agents were used to check if papers had flaws in their conclusions and it found 15% did, that kind of auditing is already a massive massive advancement for science
I and many people believe that google intentionally chooses to keep harmful ads (deceptive, scamming, or leading to malware) this is because it's often very blatant and often survives being reported
Ah I see, it was a self defensive crime against humanity and act of terror, that's good to know, say this to the children whose arms burnt off
> as for trust, avoid it We all end up running closed source code somewhere If you do you'll be much better off avoiding Israeli tech As a rule if you boycott everything Israeli it'll be smarter, more ethical, and safer
You'll have to read the license agreement for openAI or Google on their tts Or look into open source tts models with liberal commercial licenses Generally I'd stay away from the site/project/repo in the OP and just do…
It probably isn't a good tutor, I'm using frontier SOTA models and whenever I ask them to explain anything they don't consider my question and don't display any understanding but just spam out unrelated simplistic…
How does this compare with audiomass?
What do you mean
I gave it a trivial HTML job and it messed it up in several ways including being lazy and lying about results It's probably very hit or miss like everything with LLMs but I was really surprised it performed that badly
> And before anyone starts thinking about whatabouting, Yandex are based in a country where journalists are openly and publicly assasinated to intimadate Thankfully Gary Webb is here with us today to laugh at this
What's worrying for them is nothing they're doing is impossible to be recreated, if you bake it in and you don't want to pay rent you can fine tune your own model
And the only end result would be that the Chinese trained on their data just like OAI and Anthropic trained on our data so who cares
They downloaded pdfs (shock and horror) to create and publish the first open weights LLM Your ethics shouldn't be dictated by US copyright law