Fixing inequality. Properly, not the current form which overcompensates and just flips the inequality. It is a hard problem because people are unequal and shouldn't not be forced to be equal, but you want this to be controlled and not a runaway loop.

The ideal IMO is single breadwinner households with a two story house and one car and decent public transportation. Single breadwinner is particularly important - many families grow up with both parents absent for much of their lives. They don't learn what love is and this leads to broken families and crime as their children don't have a good framework.

Inequality also causes a lot of political distortion. When the rich get richer, they try to hold power. They spend more, debt goes out of control. The poor are treated like they deserve to be poor. Bad work ethic becomes the norm for this society. Increased debt and spending hits a point where the ROI is negative. The poor become increasingly desperate and murderous. Wars trigger. The empire overextends. The inability to pay debts result in bank runs. And this pattern repeats itself again and again in nearly every fallen empire.

No amount of technological advances or wealth helps if there's runaway inequality.

There is a difference between building or inventing something that really works, and getting a large percentage of the population to actually adopt it. You could say the thing you worked on 'didn't fail' if it really did what you designed it to do; even though almost no one recognized its value and put it to use improving their life, even in some small way.

History is full of very useful devices that only improved the lives of a few people; while also full of mediocre devices that were widely adopted due to very good marketing.

I strongly believe the von Neumann architecture is a premature optimization. I want to bring bit level systolic array chips into reality and democratize access to petaflops.

Existential risk from superintelligent AI, naturally. Even if it turns out to not be a risk I'd sleep a lot easier at night with an ironclad mathematical proof of why exactly it isn't on the presses. And if it is a risk, it's almost certainly the most important thing anyone could work on right now.

I'm interpreting "can't fail" as "guaranteed to succeed one way or another", here.

It's time for human-animal hybrids to become a real thing!

I'd be working on the same old things, but with a whole lot better luck :)

A personal blog that ends up becoming the biggest in the world.

Time travel.

Surfing

So what are you working on?

Ascension。

Prevent cancer?

Proving p=np

elderly care robots

Fusion Life extension FTL travel Misinformation