tedivm

The lack of OIDC support for Harbor has been the biggest annoyance for me. I'd love to be able to push from Github Actions to Harbor without needing robot users.

Depending on what they're actually pulling down this can get pretty expensive. Bandwidth isn't free.

Doing that requires write access if you're a Github Application. You can't just fork repositories back into another org, since Github Applications only have the permissions of the single organization that they work with. Rulesets that prevent direct pushes to specific branches can help here, but have to be configured for each organization.

From my understanding they were told they were no longer allowed to fundraise in the village, which meant they could not afford to continue coming.

I've spent two minutes look at this website and I still have no idea what this project is.

Once they scared off the people running the Sky Talks, which were always awesome, and messed with groups like the lockpicking folks ability to fundraise, I think the idea of it being a hacker con really died and it turned into just another corporate convention.

The current desktop client is missing support for a bunch of important things too, like signing commits.

Qwen3 has multiple variants ranging from larger (230B) than these models to significantly smaller (0.6b), with a huge number of options in between. For each of those models they also release quantized versions (your "fewer bits per parameter).

I'm still withholding judgement until I see benchmarks, but every point you tried to make regarding model size and parameter size is wrong. Qwen has more variety on every level, and performs extremely well. That's before getting into the MoE variants of the models.

It only seems like that if you haven't been following other open source efforts. Models like Qwen perform ridiculously well and do so on very restricted hardware. I'm looking forward to seeing benchmarks to see how these new open source models compare.

This isn't an occupation, it's an ethnic cleansing.

The law isn't forcing people to do this, economics are. Nothing about the law forces people to use physical books, just that they actually pay for the books instead of stealing them. The company thinks they can get away with this cheaper than negotiating for a digital copy of the book, so that's what they are doing.

> There are more than enough stories about employees complaining that they didn't get a big enough payout on an acquisition or IPO to know that this isn't true.

That's exactly why it is true. If every person who held early stage stock walked out of those events happy then no one would recommend they focus on salary.

In 2024 Tesla sales accounted for 49% of EV sales in the US. Tesla sales have been dropping quarter to quarter (ie, Q1 2025 was less than Q1 2024) as well as year over year. I see no reason why that trend would change in Tesla's favor.

I've been working in AI since 2014, and I believe "human in the loop" will be the way things have to work for awhile yet.

If it was synthetic then all this talk about it dropping the production database was purely clickbait, since there was no "production".

Honestly, the person who decided to give an LLM Agent full and unrestricted access to their production database is the person who deserves all the blame. What an absolutely silly decision. I don't even give myself unrestricted access to production databases.

It's always been the case that if you refuse to pay a bill the other party can refuse to do service with you.

No, as some portion of the $2.5b goes to Windsurf investors.

Basically, Google bought the top talent from the company. This cash was used (according to articles I read this morning) in part to pay directly out to shareholders, and in exchange Google got the top talent from the company and a license for the software (probably mostly so their new talent didn't have to worry about NDA, non-compete, and patent challenges).

Since this money went to shareholders, not to the company bank, and since top talent fleeing the company reduces the value of the company the overall value of Windsurf likely went down as part of the Google deal. This in turn likely made it cheap enough for the remainder to be purchased by Cognition.

There are sharing programs between companies, yes, but that isn't what we're talking about here.

I don't think so. The Fourth Field Army doesn't exist anymore (and hasn't since 1955). My guess is the company named their LLM initiative after this for historic reasons, and that these are more like internal project code names than anything else.

When I was at Malwarebytes we had concerns that IOBit was stealing our database and passing it off on their own. While we had a lot of obvious proof, we felt it wasn't enough for the average person to understand.

To get real proof we created a new program that only existed on a single machine, and then added a signature for that application. This way there could be no claim that they independently added something to their database, as the program was not malware and literally impossible to actually find in the wild. Once they added it to their database we made a blog post and the issue got a lot of attention.

https://forums.malwarebytes.com/topic/29681-iobit-steals-mal...

As far as I can tell xAI doesn't publish because they don't have anything to publish. It seems to me like they aren't doing anything novel or state of the art, they're just consuming other people's research and implementing it.

Sort by oldest. They had two papers in 2017 and three in 2018. It was only in 2019 that they started opening up more, but you can't change decades of culture quickly. The AI researchers I know (and I know a lot, having been VPE at multiple AI companies) would rather work with groups that are more open.

I don't think there's much mystery here. Apple has blocked their researchers from publishing and has a very siloed approach. Researchers don't like this, so they work for companies that allow them to publish and engage in the research community. As a result Apple can't hire the talent needed to execute in this space.

This was actually the approach that Vicarious AI took while I was there, and even $250m in VC funding wasn't enough to prove it out, although that may have been a problem of having too much money and not enough focus.

I think the problem (if it can be called that) is that LLMs are useful today, while we still haven't solved the embodiment problem. There's a lot more research before that'll work well, while LLMs have uses today. So the money goes to the LLMs. While it's pretty obvious that solving the problem would change society, it's also not clear how close we are to doing it. That makes it much harder to get the capital as it is a much larger risk.

Yeah, I found this article to be fascinating and there's a lot of important stuff in it. It really does feel like more people stopped at the title and missed the meat of it.

I know this is a very long article compared to a lot of things posted here, but it really is worth a thorough read.

How are you getting Qwen3:30B-A3B running with 8GB? On my system it takes 20GB of VRAM to launch it.

Honestly once Jobs died most of what Ive did was ruin existing products, such as the butterfly keyboard and the removal of all of the useful ports from their laptops.

I got hired at Kenna Security a month before they were acquired by Cisco and it was such a miserable experience that I won't work for any company the Kenna leadership are involved with, nor would I ever consider working at Cisco.

Yup, I have an umbrella policy to cover a variety of legal situations. It costs me $900 a year for a $3m (per incident) policy.