If LLMs are so good at coding…
Summary
A discussion questioning why LLMs haven't helped ROCm and Intel's software ecosystems catch up to CUDA, highlighting NVIDIA's premium pricing and the need for genuine market competition.
Similar Articles
Local LLMs aren't democratic anymore... the hardware barrier has gotten out of hand.
The author argues that running local LLMs has become inaccessible due to high hardware costs, contrasting with earlier days when consumer GPUs sufficed, and expresses frustration with the perceived lack of democratic access.
Why current LLM costs are not sustainable
The article argues that current high LLM pricing is unsustainable due to diminishing performance gains, the rise of open-weight models, specialized AI chips reducing inference costs, and zero switching costs, predicting significant price drops as competition intensifies.
What happens when LLM providers stop subsidising?
A developer shares their experience with AI inference costs after switching from subsidized OpenAI Codex to OpenRouter, prompting a discussion about the sustainability of current LLM pricing models and the potential shift towards open-source self-hosting.
Have we reached the point where open-source LLMs are “just good enough”?
A discussion on whether open-source LLMs are now 'just good enough' for most use cases, questioning the added value of proprietary models and the cost-benefit tradeoffs.
How important is it for Chinese LLMs to reach the Opus 4.8 level?
The article discusses the importance of Chinese LLMs reaching Anthropic's Opus 4.8 level, using Ramp spending data to highlight market dynamics and predicting that silicon sellers like AMD and Nvidia will be key long-term winners in the AI industry.