3 ms·
This article claims that since most current business uses of LLM's don't require high reasoning (summarization, semantic analysis), that high level reasoning wo
by atleastoptimal 3y ago
This article claims that since most current business uses of LLM's don't require high reasoning (summarization, semantic analysis), that high level reasoning won't be important for businesses, thus they would prefer smaller models, thus open source will win, etc.
This is very naive and assume that all the use cases for LLMs now will remain the same, and the enormous benefits and eventual agentic magic of higher reasoning won't soon overtake the easier LLM stuff in economic value. If OpenAI releases a model that can design, build, deploy, maintain, act as support staff for and do independent research for an entire product line of a tech company, no company is going to prefer sticking to smaller models that can summarize PDFs for them.
- ilaksh 3y agoI agree that reasoning is quite important, but I think GPT-4 is already very capable and unlocks capabilities or levels that real use-cases can leverage today. But I also think that increasing performance is just as much about curating data and model feedback, and better architecture, as it is about things like giant datasets. It looks like open source is catching up and will likely shortly reach a performance and efficiency level that rivals the current GPT-4. Open source doesn't have to be as good as the latest closed models to be useful. Once it can get a little bit smarter and more convenient then we won't need OpenAI etc. to handle many complex tasks.
- atleastoptimal 3y agoI'd agree if the scale of money being poured into closed source models was what it was only a year ago, but now it's 10x greater. Open source can't compete with hundreds of millions of dollars and the talent vacuum it creates.
- sentimentscan 3y agoYes, but then Lama from FB/meta has provided pretty good base for (open source like LLM)