Data and algorithm based architectures have led to massive leaps in AI. But what if the next step in AI is a revolution in the substrates these run on?
They've been working on 3D integrated circuit desgin for quite a long time but I'd say, more powerfully is the potential of 2.5D ICs. These can be grouped together in modular systems where it isn't one thing but a combination of many things that make it super powerful. In a way, the AI we have is already running on 3D Brains because it's running on multiple servers, stacked in racks, next to racks, and communicating with servers in completly different locations. At a Macro Level, ChatGPT is already 3D.
This speculation on 3D printing an "artificial mind" may have seemed edgy nine years ago, but in a world increasingly interested in neuromorphic computing and nanotechnology, it's looking less outlandish and more potential-packed. The idea of three-dimensional neuro-inspired compute substrates revolutionizing AI with their increased power and efficiency is intriguing, even if we're still navigating through the realms of speculation. Let's be audacious in our explorations, but also wise in considering the responsibility that comes along.
Why not just grow the brains you need and let nature take the strain on all the nanoscale architecture and, more importantly, ongoing plasticity? Reductionism isn't going to cut it. One caveat: you're going to have to give them something interesting to do...
With the convergence of technologies like AI, quantum computing, and synthetic biology happening at break-neck speed, why manufacture an artificial brain when you can grow one?
Good point! I think we're still a long way from growing biological artificial brains that have the compute abilities of silicon-based ones. There's also the challenge of being able to read/write/interrogate at speed within a 3D matrix, which would ideally mean simultaneously growing a 3D equivalent of a neural lace -- tricky with current technology! My guess is that we're going to see silicon or silicon hybrid systems first.
They've been working on 3D integrated circuit desgin for quite a long time but I'd say, more powerfully is the potential of 2.5D ICs. These can be grouped together in modular systems where it isn't one thing but a combination of many things that make it super powerful. In a way, the AI we have is already running on 3D Brains because it's running on multiple servers, stacked in racks, next to racks, and communicating with servers in completly different locations. At a Macro Level, ChatGPT is already 3D.
This speculation on 3D printing an "artificial mind" may have seemed edgy nine years ago, but in a world increasingly interested in neuromorphic computing and nanotechnology, it's looking less outlandish and more potential-packed. The idea of three-dimensional neuro-inspired compute substrates revolutionizing AI with their increased power and efficiency is intriguing, even if we're still navigating through the realms of speculation. Let's be audacious in our explorations, but also wise in considering the responsibility that comes along.
Why not just grow the brains you need and let nature take the strain on all the nanoscale architecture and, more importantly, ongoing plasticity? Reductionism isn't going to cut it. One caveat: you're going to have to give them something interesting to do...
With the convergence of technologies like AI, quantum computing, and synthetic biology happening at break-neck speed, why manufacture an artificial brain when you can grow one?
Good point! I think we're still a long way from growing biological artificial brains that have the compute abilities of silicon-based ones. There's also the challenge of being able to read/write/interrogate at speed within a 3D matrix, which would ideally mean simultaneously growing a 3D equivalent of a neural lace -- tricky with current technology! My guess is that we're going to see silicon or silicon hybrid systems first.