The author criticizes OpenAI's new o1 model, describing it as expensive, inaccurate, and unexciting. They argue that the model's "reasoning abilities" are overhyped and that it fails to deliver on basic tasks. The author also points out the legal risks associated with training these models, as they often ingest copyrighted material. They warn that if any of the ongoing lawsuits against AI companies for copyright infringement succeed, it could be disastrous for the industry.
Key takeaways:
- The author believes that the artificial intelligence boom, particularly the generative AI boom, is unsustainable and will ultimately collapse, causing significant damage to the tech industry and leading to job losses.
- OpenAI's new model, o1 (codenamed: strawberry), is criticized for its high costs, inaccuracies, and lack of new features. The author suggests that it shows OpenAI is out of ideas and desperate.
- The author also criticizes the high costs of Microsoft's AI-powered "Copilot" products, stating that they don't seem to provide much value to customers.
- There are legal concerns around the use of copyrighted material to train AI models, with lawsuits progressing against companies like Stability AI and DeviantArt. If these lawsuits prevail, it could have serious implications for companies like OpenAI, Anthropic, Google, and Meta.