Jeff Bezos aims to make a big splash with Prometheus, but what might his plan for an ‘artificial general engineer’ mean for CAD companies and the designers and engineers who use their products, asks Stephen Holmes?
Don’t worry, everyone – Jeff ’s here. Part bookshop owner, part purveyor of divorced uncle vibes, Jeff Bezos is a man richer than Midas who could buy you while you sleep and switch off a third of the internet on a whim, with what I imagine is a big red button on his desk.
Recent headlines focus on his pledge to tackle computer-aided design and it’s really put a 9-volt battery in the knickers of some of design and engineering’s most garrulous mouthpieces. But right now, most of us have more information on how to cure a hangover than we do about whatever it is Bezos is building.
The mission statement for Prometheus is to do for physical engineering what AI coding agents have done for software development
To summarise: Prometheus is his AI start-up, launched in November 2025. It has already scooped up over $18 billion in investment over two rounds of funding, with input from big-name backers including JP Morgan and BlackRock.
The mission statement for Prometheus is to do for physical engineering what AI coding agents have done for software, compressing ten-year, 100-engineer projects into one-year programmes to allow humankind to build more ‘stuff’.
In a May 2026 interview with Andrew Ross Sorkin on CNBC’s Squawk Pod, Bezos said: “What we are doing is we’re building an artificial general engineer. So, when you go to design something, we’re building tools that will make it much easier for engineers to design physical objects. So, you’ve heard of CAD… This is kind of like a very, very modern version of that.”
This is the clearest public statement of intent provided to date, and it explicitly mentions CAD, and yet we’re still no wiser as to what it might actually involve.
Behind closed doors
Finding out more about Prometheus is tricky. A quick snoop around LinkedIn suggests a team of over 100 employees, mainly software engineers who have been sourced from other large tech companies.
Based in offices in San Francisco, London and Zurich, this group’s back catalogue is mightily diverse.
From building visual effects tools to developing algorithms for drug design and mobile apps for public transport, the Prometheus crew have been involved in some pretty impressive projects. However, relatively few seem to have any background in CAD or experience of building the physical world, be it in traditional or advanced manufacturing.
Going back to Bezos’ statement, it’s worth dwelling on the bit about an “artificial general engineer”. What he seems to be describing is a type of machine learning system that can independently design and manufacture complex physical devices and is reliant on Artificial General Intelligence (AGI). In other words, it’s a computer that can think and learn like a human. The only issue is that AGIs don’t yet exist. So far, we only have what are termed ‘narrow AI’ systems, which are designed to perform highly specific, predefined tasks.
Suddenly, the idea of pumping this much money into creating an AGI simply to launch a new CAD system seems a little shaky. After all, most of the big CAD incumbents are allegedly for sale for the right price. And in recent years, swarms of AI-focused start-ups have emerged, all looking to transform design and engineering. Put simply, $18 billion would very easily fill a pretty large shopping cart with battleproven design and engineering tech.
That scenario leads to a Fantasy Footballstyle calculation in my head. Which technologies would you buy from existing vendors if your goal was to create the most advanced platform available, without having to redesign not only the wheel, but also the road to run it on?
Listening to customers
There’s also the question of what all of this might mean for existing players and the growing number of companies racing to piece AI and CAD together?
As the ‘Big Four’ CAD companies continue to drip-feed their plans for AI, we’ve seen start-ups pivot, accepting that a big vendor might quickly swallow up what they’re building. But while some start-ups will inevitably fold, many more are lining up to join the race.
Equally, the Big Four have come to realise that their big foundational models are too all-encompassing to be seriously threatened by more niche players and are increasingly happy to partner with them in order not to get left behind.
I can’t imagine a situation where a new giant built by Bezos will force CAD companies or their shareholders to radically divert course or sell up entirely. We don’t even know when a commercial Prometheus product will arrive, if ever.
What we do know is that these fastmoving times are exciting and that the emergence of AGI would certainly be transformative and could potentially leapfrog today’s design tools by some considerable distance.
But let’s not get caught up in thinking of this project as something solely focused on transforming design and engineering workflows. Instead, let’s get Katy Perry back on a space rocket. I know a guy called Jeff who could sort it out.
This article first appeared in DEVELOP3D Magazine
DEVELOP3D is a publication dedicated to product design + development, from concept to manufacture and the technologies behind it all.
To receive the physical publication or digital issue free, as well as exclusive news and offers, subscribe to DEVELOP3D Magazine here
