trentor

It's fairly priced but was not fairly advertised.

Mhm... maybe read my full comment?

I'm genuinely disappointed with my Spark. I don't know how anyone can claim it performs decently with LLMs or diffusion models. Back when I worked in VFX in the early 2000s, we had a saying: "Render time is coffee time" and if you try to run this thing with a usable context size, you'll be drinking a lot of coffee. Most of the optimizations it relies on for inference simply aren't available for training, so it crawls like a snail on almost every model. An RTX 6000 Blackwell would have been the better investment for an AI enthusiasts and for general computing there are cheaper offerings.

I like it. I used Claude and Qwen to build single file HTML decks in the last few months. The nicest thing with HTML decks is that you can not only ship the decks but integrate your prototype/demos in the same file integrated into the flow of your deck. Would love to see a feature like that.

Maybe a bit petty but I am fine with it their infrastructure really struggled in the last few months.

Local AI won't help you if an agent goes roque.

Because we are mostly discussing a single private person that got caught for maybe 20 songs. I don't want to bring up Aaron but the taste gets saltier the more we see settlements like this.

Thomas-Rasset got 80k per song and Tennenbaum got 22k per song. The law says up to 150k per work. It was a gift.

Yes, why shouldn't it?

Post hoc ergo propter hoc.

I changed all smart speakers to retrofitted old radios with an amp and a pi. The hot word detection runs on the pi itself but whisper and LLM/task orchestrations goes over my local server with a 4080.

They rode the novelty train so hard they missed that everyone is doing it better than them now.

The funny thing is they had the decision to go with some less or more pretentious...

Maybe LLMs are just Germans.

Am I missing something? I don't see a retro pixel style besides the background image?

Would love if someone with experience could chime in. I've been reading about the geo data market for over two decades now but still have no real sense of its value. What does this data typically cost? And can you specify particular locations/targeting points, or do you just get whatever's available in a broader dataset?

I hate the naming. "Open" is so abused by now that I start to distrust software that needs to explicitly state that they are "Open" because they usually aren't.

I have to admit I have every device running some sort of voice assistant on en_GB or Australian the American voices always sound like parodies or the Walmart greeter. The intonation is perpetually trying to please me, as if programmed for relentless customer service. It's hard to explain, but there's something exhausting about a voice that's always smiling.

Qualcomm is not a good software steward. Every time I used something they had their hand in, it was abandoned rather sooner than later.

Eudora: bought, milked, killed. BREW: rotted. AllJoyn: dead. Toq/Mirasol: gone in two years. CodeAurora: shut down. And the $899 Snapdragon Dev Kit: shipped months late, then cancelled with support "paused indefinitely" while units were still in transit. Even Adreno drivers barely get updates after launch.

The silicon is great. But software at Qualcomm is a launch checkbox, not a commitment. At this point, "powered by Qualcomm" on a dev platform is a signal to stay away.

You can just track the tokens used in Claude Code and codex until you hit the limit?

I like the models for creative writing. They have a distinct voice that is different from the other llms.

Same with opus nothing above medium has a reasonable improvement for the tokens spent.

You can disable automatic kernel updates on almost every distribution. Most people that use secure boot and Nvidia do it.

Runs fine on 2x4080s or on two 5060/5070s with 16GBVRAM... and faster than on the mac.

Don't ask western llms in Chinese what model they are...

Ultrasound was overlooked by US medicine as a first line imaging tool for a long time because it takes real skill and experience to do it right. But it's making a comeback. We've had Chinese, Indian, Australian, and American doctors visit us for one to two month stints to build up their skills.

Given the skill involved, it's probably a liability concern they don't want the exposure over there.