My understanding is the best answer is 'all of the above.' Nuclear is big and expensive, and just as expensive to expand, and has ongoing input costs like skilled staff, water, fuel, etc. but produces very steady, controllable power, and can do so for decades in a well designed and maintained system. Solar and wind, on the other hand, are very easily scaled down to smaller budgets, but generate intermittent power. It can be very useful to have the nuclear as a backup for moments of high demand and low solar/wind output, but very efficient to mostly use renewables because they don't have much of an ongoing input cost.

I would say that the good answer is "mox" (not sure of its english name, but it is likely similar. Used uranium is extremly dangerous for a very long time (millions of years). Yet it could be recycled and used again, and the result would be extremly dangerous for about 500 years (which is still far too much, but less crazily impossible to manage).

We already have enough used uranium to cover the next 200 years of needs, at least... And so it would be a lesser harm for future generation if we use mox than new uranium...

(But the true answer is we need new bombs, so new uranium...)

Hey, I'm in the biz. "Either or" isn't a thing. We need a completely diversified grid. Mostly nuclear would be ideal with a big chunk coming from solar/wind/hydro. Even fossil fuels have a place in certain applications.

So the planet's been using solar and wind for millions of years and it seems to have worked okay.

Wind and solar. Pop up, pop down. Nuclear needs entrenched infrastructure. The future is always renewables.

Both are needed for a healthy power structure.

Wind, solar even with batteries have low innertia (actually no innertia at all by themselves). If you want to resist turbulences on the power network without free innertia resistance by turbines (nuclear, gas, hydro), you need expensive equipment to generate that innertia other ways, and it can get quite expensive.

Hydro would be great but it's highly seasonal.

wind and solar. hydro is environmentally unfriendly but renewable. nuclear is not pratical long term, or in the short term, since it needs to build out an expensive nuclear plant, plus all the regulatory approval you need to get through, can easily take 10 years at the minimum and also there significant resistance to nuclear power plants being built in many states, plus all the need to hire expensive technicins, scientists to maintain a plant. solar, wind is much cheaper and faster to set up. Most govt is are totally resisting to fast track it. because Oil money is in politicians pockets.

our tech is not even close for things like fusion or even geothermal that is in alot of fantasy scifi genre right now.

It depends on a lot of factors, which is part of why the question is divisive. To oversimplify:

On one hand, when done optimally, nuclear provides huge amounts of power extremely cleanly and safely - even compared to renewables. There is also a number of options that could allow for significantly reduced costs and significantly increased flexibility such as SMRs, or other reactor technologies. On the other hand, nuclear has a mountain of roadblocks in front of it including antinuclear sentiment, policy blocks (particularly stemming from that sentient), and a lack of know-how increasing cost and slowing both building of reactors and development of new technologies. This means it could be amazing in the long term, particularly in regions that are suboptimal for renewables, and for industries requiring large amounts of power 24/7, but it can't do much in the short-term.

Renewables meanwhile, are cheap, fast and easy to build and easy to expand to the point where until ideal conditions, even individuals can afford to install and run their own electrical infrastructure. As such, its amazing to quickly reduce prices, reduce emissions, and reduce dependence on governments. On the other hand, most renewable options lack flexibility. Solar only runs half the day, and the further you are from the equator, the lower the efficacy. Wind takes up huge amounts of space. Hydro requires damable rivers and space to avoid destroying ecosystems. There's lots of options, but they also have large limits. As technology improves these will limits will be reduced (such as improving battery technology) but they are still present.

Ultimately, this means that where available, renewables are nearly always best in the short term. Long term, it gets messier, since it depends on what we invest in now, which technologies end up going somewhere versus being dead ends, and how the world changes in general. Most predict a mixed approach would be best, both to address the variety of global needs, and to spread risk. All that said, global warming is, at this point, a short-term problem. We can't afford to focus on the long-term anymore, which is why there is so much focus on renewables. We need to be replacing and expanding infrastructure yesterday so we don't have time to wait for nuclear, and debatably, can't afford to be investing in an energy technology that won't pay off for decades when that could be further invested in renewables.

when you say "long-run" is it human life expectancy long or heat-death of the universe long?

Wind & solar for sure. Nuclear fission is great right now, but fissile material is a finite resource. It will, like oil, eventually become too expensive to extract.

Nuclear was probably the best choice in the 1970’s; France made the best choice. But we’ve seen over and over again the problems are getting worse rather than better. Paperwork and process ever more unwieldy. Construction taking longer and longer. Radioactive waste building up with no permanent disposal. By far the highest cost power generation, and the difference is getting wider.

Much of the posts about how great it will be this time is just astroturfing by those trying to profit. No, “small modular reactors”, didn’t solve the cost, construction, or process issues, but only raised the cost of generation. No, thorium isn’t the silver bullet and really doesn’t exist yet.