40 years ago when computers took off, I witnessed all the young mathematical minds having the greatest technical potential become diverted away from natural sciences like never before.
I was already a computational pioneer myself with oils and chemicals but it was easy to see there would eventually be so many people working on advanced computer problems that I really wouldn't have to make that many of those advances myself.
So when more and more technical minds turned away from natural science technology, I doubled down on the natural science to a greater extent each year. Still not enough to even keep the still-lucrative trailing edge of technology from slipping away.
That's why I say if all you're considering is what can be done with a computer in combination with things like chemistry, biology, physics or materials, that would be limiting yourself to things that zillions of other people are getting better at all the time.
When you work with the physical materials themselves there are far more unexploited opportunities that nobody is working on because for decades too many students having the greatest potential are spending all their time on computer science instead.
The "building blocks" for a new Bell Lab aren't even available any more, they need to be "baked" for about 40 years before you could even build a foundation now.
Certainly not underrated, but I'd say RNA engineering [1]. Now that we have the theoretical knowledge of the many types of non-coding RNA (that we previously considered "junk") and the necessary biotechnological and computational tools, we can start making novel, biologically-active RNA molecules (for medicine, biotechnology, etc.)
There'll always be a demand for physical widgets with ever increasing computational capabilities. And it'lll be a while before ChatGPT can poke a circuit board with an oscilloscope and tell you that your software bug is because your I2C pull-up resistors are too large.
3d printing is underrated. It is now a very popular hobby so in that regard no. But say the tech advances to produce full colour prints and a plastic that developed that is easy to print, strong, 100c heat resistant, chemical, uv etc. You would have a mini factory in each home for many things.
I was pretty impressed with AlphaFold by DeepMind. The James P. Allison work was a big deal - they literally are curing cancer but it seems to not have penetrated popular culture for some reason.
In comp sci everything is LLMs right now, which isn't wrong from an impact standpoint, but it seems like a lot of the future value will be rigging LLMs together with a voice interface and have it "do things" with robots or agents. Theres been some discussion of the "data wall", and optimizations around the algorithms used in training, so it feels like the roller coaster is far from over. I'm personally bullish on CV/ CNN stuff which is largely overshadowed right now... Classification and predictive models still have a LOT going for them, especially since the accuracy is higher than LLMs are (99ish on discrete tasks vs ~95ish).
The rest of the comp sci stuff is interesting but still needs to prove itself as being viable. Quantum is in there for sure, but it looks like its going to be viable just over a long timeline.
Cloning hasn't been talked about in ages, I wouldn't be surprised if the first human clone either happened already or is being worked on.
The space race between the billionaires is moderately interesting when we get stuff like starlink and reusable rockets. The possibility of a moon base in my lifetime would be awesome.
I'm a huge fan of Farming robots. The space is getting some action. Precision weeding is real and being used.
There was a project I believe by the segment guys to pull CO2 out of the air, make gas, and pump it back into the ground. I thought thats great stuff.
Battery tech / Energy storage continues to be interesting. Nuclear power getting a revived interest is great. Theres stuff with molten salt and micro nuclear thats fascinating (I don't know much) Alternative storage like concrete blocks (yes its real) are also interesting.
thats all I have off the top of my head. interested to see what the answers look like
AgTech fascinates me as well - it is a space with so much potential for gain and efficiency. All the pieces for fully autonomous crop farming are available today, but just not well integrated yet. To an outsider, they don't look like hard science problems, but merely ROI and mindset changes.
To me, you're answering your own question.
40 years ago when computers took off, I witnessed all the young mathematical minds having the greatest technical potential become diverted away from natural sciences like never before.
I was already a computational pioneer myself with oils and chemicals but it was easy to see there would eventually be so many people working on advanced computer problems that I really wouldn't have to make that many of those advances myself.
So when more and more technical minds turned away from natural science technology, I doubled down on the natural science to a greater extent each year. Still not enough to even keep the still-lucrative trailing edge of technology from slipping away.
That's why I say if all you're considering is what can be done with a computer in combination with things like chemistry, biology, physics or materials, that would be limiting yourself to things that zillions of other people are getting better at all the time.
When you work with the physical materials themselves there are far more unexploited opportunities that nobody is working on because for decades too many students having the greatest potential are spending all their time on computer science instead.
The "building blocks" for a new Bell Lab aren't even available any more, they need to be "baked" for about 40 years before you could even build a foundation now.
Certainly not underrated, but I'd say RNA engineering [1]. Now that we have the theoretical knowledge of the many types of non-coding RNA (that we previously considered "junk") and the necessary biotechnological and computational tools, we can start making novel, biologically-active RNA molecules (for medicine, biotechnology, etc.)
---
[1] https://pmc.ncbi.nlm.nih.gov/articles/PMC9714562/
Embedded systems
There'll always be a demand for physical widgets with ever increasing computational capabilities. And it'lll be a while before ChatGPT can poke a circuit board with an oscilloscope and tell you that your software bug is because your I2C pull-up resistors are too large.
3d printing is underrated. It is now a very popular hobby so in that regard no. But say the tech advances to produce full colour prints and a plastic that developed that is easy to print, strong, 100c heat resistant, chemical, uv etc. You would have a mini factory in each home for many things.
I was pretty impressed with AlphaFold by DeepMind. The James P. Allison work was a big deal - they literally are curing cancer but it seems to not have penetrated popular culture for some reason.
In comp sci everything is LLMs right now, which isn't wrong from an impact standpoint, but it seems like a lot of the future value will be rigging LLMs together with a voice interface and have it "do things" with robots or agents. Theres been some discussion of the "data wall", and optimizations around the algorithms used in training, so it feels like the roller coaster is far from over. I'm personally bullish on CV/ CNN stuff which is largely overshadowed right now... Classification and predictive models still have a LOT going for them, especially since the accuracy is higher than LLMs are (99ish on discrete tasks vs ~95ish).
The rest of the comp sci stuff is interesting but still needs to prove itself as being viable. Quantum is in there for sure, but it looks like its going to be viable just over a long timeline.
Cloning hasn't been talked about in ages, I wouldn't be surprised if the first human clone either happened already or is being worked on.
The space race between the billionaires is moderately interesting when we get stuff like starlink and reusable rockets. The possibility of a moon base in my lifetime would be awesome.
I'm a huge fan of Farming robots. The space is getting some action. Precision weeding is real and being used.
There was a project I believe by the segment guys to pull CO2 out of the air, make gas, and pump it back into the ground. I thought thats great stuff.
Battery tech / Energy storage continues to be interesting. Nuclear power getting a revived interest is great. Theres stuff with molten salt and micro nuclear thats fascinating (I don't know much) Alternative storage like concrete blocks (yes its real) are also interesting.
thats all I have off the top of my head. interested to see what the answers look like
AgTech fascinates me as well - it is a space with so much potential for gain and efficiency. All the pieces for fully autonomous crop farming are available today, but just not well integrated yet. To an outsider, they don't look like hard science problems, but merely ROI and mindset changes.
Edge computing, especially when combined with AI.