tedivm

This is not satire. This is a website that has been around for years that has an email archive from emails released in lawsuits against tech companies. The page literally sites the specific case ("Tennessee v. Meta (2024)") that this came from and is easily verifiable.

Not satire. They really are just that up their own asses.

> While our company has a special role in the lives of this generation, this is likely particularly important for how I show up because I am the most well-known person of my generation.

They really are up their own asses so much in this thread. Just the arrogance of these people absolutely kills me.

Don't put words in my mouth.

When we say they're arresting US citizens I'm referring to the brown people they're picking up off the streets purely for being brown. The lawncare guy they picked up the other day, on film and with paperwork, wasn't protesting or interfering with anything.

You're just making things up to justify this fascist take over.

Before I moved to Chicago I used to go to RSNA for work. If I didn't live in Chicago already I wouldn't be traveling here.

I live here, and Pritzker says the right things then does the wrong things. He sent the Illinois State Police to help protect ICE from protestors as Broadview, freeing up their resources to attack and kidnap people. In half the videos you see out there his state police as assisting ICE. We have Chicago Aldermen out in the streets and in the hospitals getting arrested and assaulted, we have candidates for congress like Kat Abughazaleh being indicted for protesting, and then Pritzker is giving speeches and going on podcasts while not even stopping his own thugs from helping ICE.

He doesn't have a spine, he has an election strategy.

They unflagged it, but the pause in upvotes means it dropped off the front page and out of people's minds.

So what you're saying is that the rate of violence has been dropping over the last decade? And that we should ignore fed violence (against brown, white, and black people)?

There's a lot of good independent media looking at this. Some good sources:

https://bsky.app/profile/unraveledpress.com

https://bsky.app/profile/djbyrnes1.bsky.social

https://bsky.app/profile/thetriibe.com

https://bsky.app/profile/prisonculture.bsky.social

Until the supreme court overturns that.

I live in Chicago, and this article doesn't even scratch the surface of how bad it is. My wife went to the beach yesterday for 10 minutes to try and rest from the chaos and a fucking black hawk helicopter buzzed by her. They literally fly over my house daily.

People, US citizens included, are literally being abducted. People have been shot and killed by masked agents. People have had their children abandoned on the side of the road after being kidnapped. Just today they raided Little Village with hundreds of masked troops. I'm in a dozen signal groups to get alerted about where things are.

What scares me the most is how few people seem to actually know what is happening here. I talk to people outside of Chicago, and watch the news, and I don't see or hear about anything that's going on here. I tell them what's happening and they are shocked.

It is impossible to convey what is happening here, how scared we all are for this country, and how much things seem to escalate every single day that this goes on.

Edit: This post has been flagged and hidden, just demonstrating how much this country wants to pretend this isn't happening. It's unflagged now, but the fact that anyone would want to hide what's happening here shows how bad things are for all of us.

Using UTC for servers was standard when I entered the field in 2005.

This is just so inaccurate, at least for California.

It goes beyond growth though, and directly to how they treat their staff.

If you actually read the specific bug and use of a private method it really was a stupid decision by one developer awhile ago that just fell through the cracks. There really wasn't a benefit to doing what they did, which is why their fix was to just go back to using public APIs.

I think the failures here are that Apple should have tested this themselves and the Electron devs should have tested and resolved this during the beta period.

For the less affluent you can setup a Jellyfin media server and rip your own blu-rays with makemkv.

Yup- ESP32 is absolutely compatible with Arduino.

There's also great support for CircuitPython and MicroPython, which makes it trivial to program the devices.

It really depends on what you mean by RAG. If you take the acronym at face value yeah.

However, RAG has been used as a stand in for a specific design pattern where you retrieve data at the start of a conversation or request and then inject that into the request. This simple pattern has benefits compared to just using sending a prompt by itself.

The point the author is trying to make is that this pattern kind of sucks compared to Agentic Search, where instead of shoving a bunch of extra context in at the start you give the model the ability to pull context in as needed. By switching from a "push" to a "pull" pattern, we allow the model to augment and clarify the queries it's making as it goes through a task which in turn gives the model better data to work with (and thus better results).

I've been working with Python for years (since 2014) and typing makes the code less buggy and easier to maintain. I also would hardly call it "shoehorning", as years of design went into it.

Honestly the only people I see who really push back against it are the people who haven't bothered learning it. Once people use it for a bit, in my experience at least, they don't want to go back.

The thing that is interesting though is that China is releasing the models they train (for the most part). Unlike OpenAI, for instance, China seems to still be developing in an open way including sharing code, weights, and publishing techniques.

These models are pretty amazing too. Their performance remains high even with smaller models, allowing them to competitively run on consumer grade hardware. I can (and do) run the Qwen models at home and use them for real tasks. As the tool ecosystem matures we can offload what the model is bad at to other systems, which makes it even easier to use smaller models. The focus on efficiency and performance from the Chinese companies has had huge practical gains.

Any US startup, or even large company, in the AI space can and should take advantage of this and use these models. The cost benefit compared to something like Azure or OpenAI is massive.

There are people who have had to move after data breaches exposed their addresses to their stalkers. There's also people who may be gay but live in authoritarian places where this knowledge could kill them. It's pretty easy to see a path to lethality from a data breach.

I actually am not using the 3090s as one unit. I have Qwen3-30B-A3B as my primary model and it fits on a single GPU, then I have all the TTS/STT on the other GPU.

I'm actually using the esp32 s3 boxes myself!

For the physical hardware I use the esp32-s3-box[1]. The esphome[2] suite has firmware you can flash to make the device work with HomeAssistant automatically. I have an esphome profile[3] I use, but I'm considering switching to this[4] profile instead.

For the actual AI, I basically set up three docker containers: one for speech to text[5], one for text to speech[6], and then ollama[7] for the actual AI. After that it's just a matter of pointing HomeAssistant at the various services, as it has built in support for all of these things.

1. https://www.adafruit.com/product/5835

2. https://esphome.io/

3. https://gist.github.com/tedivm/2217cead94cb41edb2b50792a8bea...

4. https://github.com/BigBobbas/ESP32-S3-Box3-Custom-ESPHome/

5. https://github.com/rhasspy/wyoming-faster-whisper

6. https://github.com/rhasspy/wyoming-piper

7. https://ollama.com/

This is exactly what I do. I have two 3090s at home, with Qwen3 on it. This is tied into my Home Assistant install, and I use esp32 devices as voice satellites. It works shockingly well.

It feels like that's the direction most people are going in, but that doesn't change the fact that no one is going to trust Broadcom again after this.

I have two 3090s, but I'm only using one for this model.

This is only if you ignore the growing open source models. I'm running Qwen3-30B at home and it works great for most of the use cases I have. I think we're going to find that the optimizations coming from companies out of China are going to continue making local LLMs easier for folks to run.

The whole reason to test things is because these types of systems will behave in ways people didn't expect. While people may not have been able to guess that it would turn out exactly this way, that's the whole reason why they should have actually tested for unknown consequences.

Instead it sounds like they rushed to release this as quickly as possible, skipping all sorts of testing, and people died as a result.

That's a bit of a stretch. Lots of stuff that gets published still has patents associated with it. Just because something is done in the open doesn't mean it can't be commercialized.

You're mixing up Human OIDC and Machine Flow OIDC. You can use OIDC to log in as a user, but you can't use OIDC to allow federated trust from something like Github Actions.

If you can find an example of OIDC with Github Actions and Harbor I'd love to see it.