Hacker News Reader

Comment by wileydragonfly | Hacker News Reader

I always wonder where we’re going to see the next kind of world changing innovation like this.

Immunotherapy? Maybe.

Recently all I’ve gotten is a phone that listens to my conversations and serves me targeted ads.

Feels like a lot of people who should know have been saying biotech for a while, but I’m unclear what big breaks are even being worked on. Not that targeted cancer therapies isn’t super cool and worthwhile work, but it does feel a successful result still basically fits the existing industry shape; it’s not the kind of thing that would be spawning entire new industries the way computers and the internet did.

I think a big part of that comes down to regulation when it comes to biotech, which, frankly, I think is a good thing. After all, skimping on software (usually) won’t kill anyone, but imagine how much worse things could have been with something like Theranos absent regulation.¹

⸻

1. For those rushing to point out that regulation didn’t stop Theranos, it might not have kept the fraud from starting, but it did keep it from continuing.

Pointing out Theranos would be like pointing out the existence of double agents despite rigorous counter-measures. Things will happen; one wants them to happen less.

The groundwork isn't there. Computers were the result of blending Maxwell, Boolean algebra, and some fairly obscure speculations about computability in the most abstract sense.

There's nothing equivalent waiting for someone to industrialise it. AI is being hammered together out of bits and pieces, but there's no Grand Theory. Robotics is the same, but further behind. Bio is similar but maybe a little ahead.

Physics and math just haven't primed the pump for another revolution.

How do you know? The pump could be primed with all sorts of obscure, non-obvious things the importance of which only becomes evident after the fact when someone combines them.

I think the importance of Boolean algebra and computability theory to early computers is vastly overstated. I see no influence of computability theory on ENIAC, for instance. As for Boolean algebra, it seems more descriptive than useful for early computers. But I'm very interested if people have counterexamples.

Millions of people, at all stages of the age curve, not dying or losing their marbles to diseases that can be treated by immunotherapy? I foresee many opportunities.

Time will tell…

As an aside, immunotherapy is pretty neat stuff. My fathers bladder cancer is in remission thanks to it.

Yes, there are still some issues being worked out (possibility to cause diabetes) but it is just neat to see in action. The medical field is one that still seems to be making these kinds of leaps. It is wild seeing some diseases that only a decade ago were a death sentence or had extreme solutions now becoming treatable.

I think that we’re definitely seeing some world-changing things happening in the vaccine world with mRNA vaccines. Covid accelerated development, but there was a lot of movement before that already.

LLMs no good for you?

Not OP, but what game changing uses are you finding for LLMs? I mostly find they're good for brainstorming ideas of activities to do, asking "is there a thing like this and what is it called", and writing trite correspondence to acquaintances. It's useful for me but hardly life changing.

Putting it like that, and I like putting it like that, sounds like LLMs are to Wikipedia what shoes are to people who want to go from A to B.

That computers can reasonably handle natural language instructions is pretty amazing.

Reply on news.ycombinator.com