I worked on my master's thesis quite a bit, even worked on submitting a publication (it was in control systems and was only rejected because they wanted more empirical results). The stuff was extremely groundbreaking - we were able to get 2 orders of magnitude of better performance compared to existing solutions.
Fast forward to a few weeks back, when I was talking with a process team at a multi-billion dollar biotech company (yes, it's a huge) somewhat casually, when that idea cropped back up. Their response was basically "meh, we really don't need that much higher performance".
I'm sure if I were to start up around this, I would be able to get enough funding from the recent AI rounds from VCs. But I wouldn't have customers willing to buy this in the first place! I would be like the proverbial solution looking for a problem to solve.
I worked on my master's thesis quite a bit, even worked on submitting a publication (it was in control systems and was only rejected because they wanted more empirical results). The stuff was extremely groundbreaking - we were able to get 2 orders of magnitude of better performance compared to existing solutions.
Fast forward to a few weeks back, when I was talking with a process team at a multi-billion dollar biotech company (yes, it's a huge) somewhat casually, when that idea cropped back up. Their response was basically "meh, we really don't need that much higher performance".
I'm sure if I were to start up around this, I would be able to get enough funding from the recent AI rounds from VCs. But I wouldn't have customers willing to buy this in the first place! I would be like the proverbial solution looking for a problem to solve.