Right now, the capabilities of open source AI models like Meta's Llama are probably 2 to 3 quarters behind more closed LLMs like OpenAI's GPT-4, says Sandeep Bhadra — an eternity in the fast-moving AI market. But in the long run, he argues, open source AI has the potential to not only catch up, but get ahead. It's especially good news for those startups who are willing to take on the challenge of fine-tuning those open models on an enterprise customer's own data, going deep on specific use-cases. Watch more here: #ai #artificialintelligence #opensource #startups #startupadvice
Jean Schmitt
5mo
Well said !
VP Product @ Attentive | AI Professor @ UCLA
5moThis is right on. It's going to be a very interesting world when the public cloud models start to plateau and the world moves to customized open source models. These can capture the best of both worlds - the latest in general models (ie human speech) with narrowly trained use-case specific models (ie product recommendations).