makin

Companies are collections of people, and these companies keep losing key developers to the others, I think this is why the clusters happen. OpenAI is now resorting to giving million dollar bonuses to every employee just to try to keep them long term.

I was a bit confused as to what "Cognition" was, but they're the makers of Devin (edit: that just got added to the title, for reference), so that makes sense. Just buying the competition, the only surprise is they had more money to spend than the big ones.

They already do this, it's the moderation model.[1]

This name thing is an additional layer on top of that, maybe because training the model from zero per name (or fine tuning the system message to include an increasingly big list of names that it could leak) is not very practical.

[1] https://platform.openai.com/docs/guides/moderation/overview

I think it's a bit absurd to care about trademark infringement of a social activity like this, but it should have been "Advent of AI Code" at least.

Personally I didn't find the movie boring overall, but there were around five too many romance scenes between Driver and Emmanuel's characters that didn't seem to move anything forward, the kind of scenes that usually get cut for redundancy.

I'm sorry if this is strawmanning you, but I feel you're basically saying it's in the public's interest to give more power to Intellectual Property law, which historically hasn't worked out so well for the public.

The main issue was that Meta quickly took down the first announcement, and the only remaining working submission was the information-sparse HuggingFace link. By the time the other links were back up, it was too late. Perfect opportunity for a rescue.

Once the data is "compressed" into the model it cannot be easily removed without starting the training over.

A bit of a shame about the exploit applying to THUG PRO. The mod is played to this day, since the more competitive side of the Tony Hawk franchise has been dead for almost twenty years (with the exception of the THPS1+2 remake, which was but a blip in the scene).

The mod itself is over 10 years old now, and I think the original developers are gone, explaining why no one was interested in fixing it when Ryan reported it. But this means that now the mod is unusable, no one is going to want to risk a full privilege exploit taking over their PC.

Hopefully this article reaches someone who's a bit more interested in patching the mod.

If you check the issues, you'll learn this is not a supported project anymore (and honestly, it hardly worked even back then).

Two big issues with Archive.org are that 1. it's a single point of failure, they don't encourage mirror sites to emerge, and 2. they keep using the "brand" to fight unwinnable battles like hosting books they don't own online, risking the whole endeavor.

I still appreciate it, but just imagine if it goes down due to a lawsuit. Now that Google no longer shows cached results, an entire historical record would be gone.

I think quality is fine, the issue is the strings, having "Legendary Wire" in a serious factory game.

Denuvo mostly works. Allegedly they have a custom approach to each new game, so cracks can take months to appear, with some unpopular games never having been cracked at all. The price is lowered performance, of course.

This is amazing, but surely analyzing the Google car itself would be against the spirit of the game for the hardcore players? Is it a standard tactic?

> We recognize that generating speech that resembles people's voices has serious risks, which are especially top of mind in an election year. We are engaging with U.S. and international partners from across government, media, entertainment, education, civil society and beyond to ensure we are incorporating their feedback as we build.

People had been theorizing the slowdown in releases was because of this, I didn't expect they would come out and admit it.

I've found the quality of Elevenlabs' voice generation is at least twice as good as this, which is surprising since OpenAI has had more time and money to work on this. There seemed to be an issue with inconsistent accents in the multilingual samples.

That final boss, Trimming the Herbs, has definitely turned out to be a hacked (TAS) level, so The Last Dance was the last, already beaten level. This article is fairly outdated for a developing story.

Personally I find it really useful for juggling interpreters between projects and keeping their state alive. There are CLI solutions for this, of course, but it's another set of commands to remember.

Aren't we also machines? If we make theories that discount our own intelligence that's just throwing our hands up and giving up.

Maybe I would agree with you if LLMs weren't already embedded in a lot of useful products. I myself measurably save a lot of time using ChatGPT in such "unintelligent conversations". It's intelligent in the ways that matter for a tool.

There isn't much incentive in making new stuff when it can be shut down by Sony or whatever company at any moment. See A2MR.

Technically we can, that's what negative prompting[1] is about. For whatever reason, OpenAI has never exposed this capability in its image models, so it remains an open source exclusive.

[1] https://stable-diffusion-art.com/how-to-use-negative-prompts...

I do notice that she's written around some of the technical things she's lost with the move: footnotes are no longer clickable, and there is no dark mode. Nice article regardless, the Github repos are appreciated.

You have some performance comparisons, but I'd like to see a table of the features it has over the mainstream frameworks. Is everything in the list at the top of the article something neither Flask nor FastAPI offer?

The parent comment was referring to free ($0 a month) models. With those, companies don't need to plan around the possibility that Stability AI hikes up the price afterwards.

There is a theory that GPT-4's secret sauce is a combination of commissioned high quality training data (which they don't really hide) and an unknown implementation of Mixture-of-Experts [1].

No proof, of course, as the other comments have said, they won't share. The Open AI name is still ironic. I've also still not seen another successful attempt of MoE by any other company, which you would expect to if it was true.

[1] https://hkaift.com/the-next-llms-development-mixture-of-expe...

Keep in mind this is 80% of the subreddits that announced would be participating in the event. There are over 3.5 million subreddits altogether, though of course most of them are dead or with a single user.

I think Archive Of Our Own (AO3) solves a much easier problem (hosting bbcode text), and even then, after many years, they are unable to get enough volunteers to implement a direct message system. It's not quite apples and oranges, but I still don't think the same model could possibly work for Reddit. Yes, Reddit users have made some amazing tools for moderation, but they are still fully dependent on preexisting, complex engineering work by the Reddit employees, including the API. Maybe a better comparison would be Wikipedia, but again, much simpler problem.

I don't think there's a real answer anywhere, but my guess is that it's not about selling the product per se. They want to introduce the concept to people, get them used to it, so non-owners don't cause a wave of bullying like what happened with Google Glass, whose users came to be known as "glassholes" [1].

[1] https://www.wired.com/2013/12/glasshole/

I think Tildes is pretty good, but their subreddit-equivalent policy is bizarre—only admins can make them, and they will only make them if they see demand. This is a chicken-and-egg problem that will prevent many niche topics from finding a community, the way they did on Reddit. I hope it's just marketing talk covering a technical limitation instead of a genuinely permanent flaw.