Ten years from now… Will OpenAI be a major force in the economy? I’m not sure. Will Google be the go-to firm for AI (however AI is defined in the future)? Maybe.
On Scott Galloway’s podcast, he mentioned the idea of investing in ETFs (around the 7:50 minute mark). To avoid downsides in investing, instead of looking for a needle in the haystack, buy the whole haystack.
We should take a similar approach when it comes to AI. At the moment, large language models are all trained on huge, overlapping sets of data. Firms are fine-tuning based on specific prompts, industry considerations, and/or proprietary datasets. It’s not clear at the moment who will be the needle in the hAIstack. Or if that needle even exists yet. But the big idea, of computation surfacing previously unseen patterns because of the scale and speed at which it can parse data, that’s not going anywhere.
It’s a great time for experimentation and learning. As these AI companies figure out their business models, you can test drive them, learn more about what they are and are not good at, and develop a more refined taste for how to work with AI. Then, as the field becomes more clear, you’re ready with your well-defined use cases and experience in the space to leverage more mature business models for your specific needs.