Our investment thesis around the extraordinary recent progress in AI due to developments in new foundation models is that the opportunities are not in AI per se (unless you are a large technology platform already) but in the application layer on top, and in particular in sophisticated user interface design. The background and reasons for this are outlined below.
AI, post ChatGPT
Many people assume technology investment is primarily around innovation in hardware and software, but in many ways it is more about investment in here-and-now business opportunities that make use of cheap, commoditized technology, particularly when these leverage the benefits of networks that the internet allows.
The history of the internet has been one of commoditization of networking through open protocols and message standards (where one of the most popular of these, RSS was co-authored by one of our advisors). Open source applications for operating systems, application servers and database management systems, commoditized back end software and on top of this, cloud computing commoditized server hosting and sysadmin requirements.
Recent developments in AI have resulted in large foundation models which create general purpose tools to allow for auto generated front end application development and individual actions or use cases (e.g. asking for the sky to be removed from an image).
This commoditizes the final layer of software services – application development – and opens up a whole raft of new opportunities for low cost, interfaces from curated ‘prompts’. Niche, highly specific enterprise software applications are now economically feasible and consumer products based on ‘semantic interfaces (a single, ‘remove sky’ button rather than a dozen photoshop mask editing tools) will allow many tasks from legal contract creation to web design to be accessible to anyone.
The Wood from the Trees
The problem with capturing opportunities in AI vs previous significant advances is that the ease of entry means that the space is incredibly crowded.
The best emerging opportunities will likely cluster around specific ‘scenes’, either geographically distributed like 3D rendering and games or around particular centers of focus, such as happened with fintech. Our strategy is to identify these clusters and then drill down for the opportunities within them.
Aperture | Hub
Sharing the latest Insights on embedded finance, digital assets, fintech, go-to-market, venture capital and more.