I make computers
the Chinese AI lab also released a smaller, “distilled” version of its new R1, DeepSeek-R1-0528-Qwen3-8B, that DeepSeek claims beats comparably sized models on certain benchmarks
Most models come in 1B, 7-8B, 12-14B, and 27+B parameter variants. According to the docs, they benchmarked the 8B model using an NVIDIA H20 (96 GB VRAM) and got between 144-1198 tokens/sec. Most consumer GPUs probably aren’t going to be able to keep up with
On one hand, it’s inane how hard Anthropic is trying to anthropomorphize Claude with these experiments and scenarios. It’s still just a chatbot. On the other hand, as these products inch closer to demonstrating true intelligence, we’ll be glad someone was at least thinking about the implications during the early stages of development.
You can easily sync your personal music collection to your iOS device using the macOS “Music” app in tandem with the Finder, or using iTunes on Windows. I’ve not explored the options on Linux, but I suspect they’re out there.
I’ve got a personal collection that’s growing steadily, mostly from CDs and digital purchases. I do not use steaming services, and my iPhone is my primarily listening device.
I am generally opposed to the integration of generative AI in consumer hardware, since it doesn’t have much practical utility at this point.
However, the features described in this article mostly have to do with extracting information from images. This is actually quite useful! For example, macOS allows users to select text and automatically mask objects from images. It’s a feature I use heavily and wish other operating systems had good support for.
In a world where unfettered Internet access has completely eroded our ability to form connections with others, the solution to loneliness couldn’t possibly be anything but more screen time 🤦🏽♂️
On my Mac mini running LM Studio, it managed 1702 tokens at 17.19 tok/sec and thought for 1 minute. If accurate, high-performance models were more able to run on consumer hardware, I would use my 3060 as a dedicated inference device