Comment on AI haters be like:
vapeloki@lemmy.world 1 day agoPlatform and systems Architect for a german managed Services provider here.
I second that. Currently, I am working on replacing external AI providers with 150W AMD hardware. Most of our teams used LLMs for stuff where classical ML is far more reliable and faster.
The few places where LLMs are needed, we still can do locally.
So, instead of spending millions in some tokens that take gigawatt of Power, we estimate that our 5k employees will need a total power of 5kw across Europe. We have 100x the solar distributed across our locations.
We conflate LLM or Diffusion with AI here, because most had never contact with other kinds of AI.
EffortlessGrace@piefed.social 1 day ago
I began my DNN career path researching cancerous tissue for medical diagnostic imaging which is a computer vision problem. I think it would surprise a lot of people how power efficient you can make DNNs when it comes to other tasks besides predicting text or generating media. It pleases me greatly to see other peers in my field share my prognosis about language models compared to other DNN algorithms for different modalities
vapeloki@lemmy.world 1 day ago
Maybe this helps for people stumbling over this thread: A Google Coral device can do classification on images with around 18fps. That is a 2w USB device.
That is the power of a well built DNN.
And you can adapt this this stuff like Monitoring data, network traffic and much more. All “AI”, same basic technology like ChatGPT and Claude but made for specific tasks. And they do their tasks damn fucking well