Computer-assisted mathematics and autonomous weapons though are extensions of things that already existed without AI, and the application of AI to warfare in particular is fraught with potential problems; what happens when say AI misidentifies a target and kills friendlies or civilians for instance.

AI has thus far proven problematic in almost every area of society it's been shoehorned into; it can certainly be useful in some cases, but its pitfalls and limitations are numerous.

In terms of profitability, if they solve quantum AI then it will be worth hundreds of trillions at the least and justify all the AI expenses by itself. I know people don't want to hear that but too bad. Quantum is the real prize at the end of the rainbow for mega-corporations and governments ... and possibly the biggest disruption to modern society that we will ever experience. Also, thank you Trump for making my quantum stock jump by almost 50%.

Before Quantum computers are figured out, I think we'll have pretty intricate analog and digital photonic processors, which are more immediately suited for data processing and deep learning than quantum computing

That'd get us about three orders of magnitude more energy efficiency per compute unit. 

I have a friend who is helping extend Lean (the proof verification language used in the Erdos problem solver that has been recently in the news) to have enough mathematical foundations to formalize quantum mechanics,with automated photonics research as the ultimate end-goal (and where his funding comes from.)

Photonics has already transformed networking, so it makes sense that it will extend to specialized processing. 

Of course Quantum Computing will help with the specific class of intractable problems that classical computers can't solve, and that will indeed be revolutionary. 

My main concern is society and how it will handle this. "Corporations" and "states" might become outmoded social institutions. 

https://isaiprofitable.com/

What's going on right now is the beginning of a truly massive shift though, like just last month Ukraine captured a Russian position using only ground robots and aerial drones for the first time so no infantry were needed. They're still mostly human controlled but every year autonomous capabilities are increasing so it'll get to a point where if a country doesn't care if civilians are harmed it could just send in swarms of drones and robots to cause mass destruction.

There are both significant downside and upsides currently. More than anything what I'm getting at is the technology is already powerful enough to do a lot of good and bad things. It's gotten to the point where a person can get a lot of utility out of it in positive ways to do good while another person can get a lot of utility out of it in negative ways to do bad.

I remember 5 years ago, during the pandemic, there were many predictions that there would be a collapse in delivery apps, because all of the delivery apps weren't making profits. Many industry analysts suggested that these companies would never become profitable and that it was impossible given the high operational costs.

Roughly around 2023-2024 they became profitable. 

This is a common trend in the tech industry. Companies make large losses over time, and then a decade or two out they start making large profits. 

Compute costs are going to decrease over time as infrastructure is built out and currently nascent technologies (high bandwidth memory and networking, photonics, etc) mature. 

We need to come up with a way to have AI without data centres and without the consequences of today's AI, like misinformation, job losses or deepfake child porn (thanks, Elon). If we really can't completely get rid of AI, we should at least fix it so it can benefit everyone.

NVidia *lmao*

I remember the dot.com crash and a certain company that had always written losses up to that point, even increasing them. But then, in the crisis they started making winnings. That was Amazon. I think that is the pattern: we will have quite a lot of companies that go bankrupt or are bought out, but a few winners.

And yes, one way or another profitability will come. Either by cost reduction or by milking customers (preferably company customers)... or both. At which point we level out is the question though, what will the tech look like?

Hopefully taking up way less space and not stealing our drinking water or jobs.