Artificial Intelligence as of late November 2022, the date of launching ChatGPT to the public has caught everyone by surprise. The big bang of deep learning that started off in 2012 from a computer vision competition in Toronto, is no longer just a method for scoring higher accuracies on a large image dataset.

AI has now become a central part of a lot of modern technologies, and the trend is that it will continue to take over more parts of the GDP of modern economies.

It’s been quite known for a while in the startup world that AI for every vertical will likely be a successful business, and that AI “Copilot for X” is now on the front page of many business pitches. Microsoft for example recently announced a lot of exciting integrations of copilot capabilities into their Teams products like Powerpoint and Excel.

As a programmer myself I’ve benefited a lot from copilot for code, outsourcing the repetitive non-creative tasks of software development to a simple tab from an auto-complete engine.

One application that has become really popular in the software world is AutoGPT, and I think we can expect a lot more AutoGPTs for different verticals just like Copilot.

What’s different about AutoGPT is that it’s not only an AI assistant, but rather that it can actually produce economic value autonomously all by itself. The implications of such systems in production are yet to be discovered, but we can expect to see more business value provided by such models.

The only blocker that I can think of in adapting AutoGPTs in the real world is the reliability of these models at this moment, for example, Tesla Autopilot is great in navigation yet still faces some reliability problems in the real world where humans can’t yet 100% rely on it for driving.

Exciting times ahead!