Prometheus Industries is a company that might become very important to additive manufacturing in the future.
What is Prometheus? It is a startup company launched by Amazon’s Jeff Bezos late last year. The goal is to develop an AI tool that provides engineering and manufacturing, which is quite different from today’s AI tools that generally work on digital data: text, images, video, and audio.
Apparently, they will build an “artificial general engineer” that can not only design complex physical systems but follow through with prototyping and even manufacturing. Supposedly, the target domains include computers, automobiles, spacecraft, aerospace, medical devices, jet engines, drug compounds, and consumer electronics, which is quite a broad list.
Bezos’ goal here is to shorten the “idea to product” cycle, which today can be quite lengthy, particularly on highly complex projects. Steps likely include design iteration, prototyping, simulation, manufacturing process planning, and probably testing feedback loops.
The biggest initial challenge for Prometheus will be training data. The chatbots of the world were trained on a planetary internet full of text, audio, video, and images. There is really no such equivalent for product design data.
There certainly are CAD files, but a large amount of design still remains in 2D blueprints and even on paper. Formats are all different and generally inaccessible due to their proprietary nature. It’s unclear how Prometheus will gain access to training data, but I have an idea we’ll discuss in a moment.
It’s been rumoured that Bezos intends to acquire a series of larger manufacturing companies to serve as a base for conversion to AI-powered processes. This would make sense, as they would provide a working platform to deploy and experiment with this entirely new concept.
Prometheus, through Bezos, has raised significant funding. According to Crunchbase, the company has received a staggering US$18.2B — so far.
With that amount of cash, Prometheus could easily pay a series of manufacturers for access to their manufacturing data for training purposes. Not all would participate due to regulatory constraints, but with suitable terms and a fat payout, I suspect many companies will take part.
Assuming this concept works, you can immediately see how it would benefit the 3D print world: at some point in the lifecycle, 3D printers (and other equipment) would be used to produce prototypes and even final end-use products.
If Prometheus can speed up the idea-to-product lifecycle, then there should be a corresponding increase in the use of manufacturing equipment, including 3D printers. We can all see how existing AI tools have grown the capacity of many roles in business, so there’s an expectation that would also occur in this area.
So far there’s not much more to know, as the company is quite secretive. In fact, they don’t even seem to have any content on their basically blank website. There’s not even a logo, but there is a privacy policy.
Most information about the company is found in a few media interviews:
While the 3D print world would be significantly affected by a successful Prometheus, one can only imagine how the rest of the world might change. When new products appear ten or a hundred times faster than they do today.
Via Prometheus and Ars Technica
