You are probably confusing fine tuning with training. You can fine tune an existing model to produce more output in line with sample images, essentially embedding a default “style” into every thing it produces afterwards (Eg. LoRAs). That can be done with such a small image size, but it still requires the full model that was trained on likely billions of images.

deleted by creator

While I agree that the AI they will implement will likely not be very effective, it doesn’t have to be to cause massive human suffering. Eg. Google incorrectly marking exposed photos of your kid for your doctor as CSAM. There’s also no guarantee that once these companies finally wake the fuck up (If they’re not already completely aware what they’re doing is messed up) that they will close these holes they’re punching, and that could mean they could replace AI with a mass surveillance tool at any point without you knowing. Nobody should be a fan of this.

I’m sure such cases exist, but where I’m from people don’t really get paid to host turbines, maybe companies at times. They dislike them because it affects the view in the area, and especially if you live very close to them the blades can cause noticeable flickering shadows. That latter point has a lot more weight to it in my eyes, but people do really care about the former as well, and it’s kind of hard to push on people when they live there and not you.

There is competition in battery production. Pretty much all of society would be better off with better batteries, so price gauging in an industry like that is quite hard. And if it was, it would not go unnoticed.

The problem is simply the technology. There’s advancements like molten salt batteries, but it’s practically in it’s infancy. The moment a technology like that would become a big improvement over the norm, it would pretty much immediately cause a paradigm shift in energy production and every company would want a piece of the pie. So you’ll know it when you see it. But it might also just start off very underwhelmingly like nuclear fusion and very gradually improve with the hope it can scale beyond the current best technologies for batteries.

All we can do is wait and hope for breakthrough, I guess. Because cheap and abundant batteries could really help massively with reducing our carbon output.

2.160 GW is it’s rated capacity. I’m not sure how you got from there to 14.2 dollars per watt, but it completely ignores the lifetime of the power plant.

Vogtle 3&4 are really a bad example because unit 4 only entered commercial activity this year. But fine, we can look at what it produces just recently.. About 3335000 MWh per month, or about 107 GWh per day. We can then subtract the baseline from Reactor 1 & 2 from before Reactor 3 was opened, removing about 1700000 MWh per month. Which gives us about 53 GWh per day. The lifetime of them is expected to be around 60 to 80 year, but lets take 60. That’s about 1177200 GWh over it’s lifetime, divided by the 36 billion that it cost to built… Gives you about 0.03 dollars per kWh. Which is pretty much as good as renewables get as well. But of course, this ignores maintenance, but that’s hard to calculate for solar panels as well. As such it will be somewhat larger than 0.03, I will admit.

Solar panels on the other hand, often have a lifetime of 30 years, so even though it costs less per watt, MW, or GW, it also produces less over time. For solar, and wind, that’s about the same.. So this doesn’t really say much.

But that wasn’t even the point of my message. As I said, I agree that Nuclear is slightly more expensive than renewables. But there are other costs associated with renewables that aren’t expressed well in monetary value for their units alone. Infrastructure, space, approval, experts to maintain it.

Let’s ignore that no grid in the country actually needs 10hr storage yet.

Because they cannot. They can’t do it because there’s not enough capacity. If the sun is cloudy for a day, and the wind doesn’t run. Who’s going to power the grid for a day? That’s right. Mostly coal and gas. That’s the point. Nuclear is there to ensure we don’t go back to fossils when we want to be carbon neutral, which means no output. If you are carbon neutral only when the weather is perfect for renewables, then you’re not really carbon neutral and still would have to produce a ton of pollution at times.

I’m glad batteries and all are getting cheaper. They are definitely needed, also for nuclear. But you must also be aware of just how damn dirty they are to produce. The minerals required produce them are rare, and expensive. Wind power also kills people that need to maintain it. Things aren’t so black and white.

Also consider that PV and batteries have always gotten cheaper over time, while nuclear has always gotten more expensive.

This is not true, and it should be obvious when you think about it. Since this data fluctuates all the time. Nuclear has been more expensive in the past, before getting cheaper, and now getting more expensive again. Solar and wind have had peaks of being far more expensive than before. These numbers are just a representation of aggregate data, and they often leave out nuance like renewables being favored by regulations and subsidies. They are in part a manifestation of the resistance to nuclear. Unlike renewables, there are many more steps to be made for efficiency in nuclear. Most development has (justifiably) been focused on safety so far, as with solar and wind and batteries we can look away from the slave labor on the other side of the world to produce the rare earth metals needed for it. There is no free lunch in this world.

For what it’s purpose should be, which is to provide a baseline production of electricity when renewables are not as effective. A higher price can be justified. It’s not meant to replace renewables altogether. Because if renewables can’t produce clean energy, their price might as well be infinitely high in that moment, which leaves our only options to be fossil fuels, hydro, batteries, or nuclear. Fossil fuels should be obvious, not everyone has hydro (let alone enough), batteries don’t have the capacity or numbers at the scale required (for the foreseeable future), and nuclear is here right now.

Solar and wind are cheaper yes. Batteries, no. If batteries were that cheap and easy to place we’d have solved energy a long time ago. Currently batteries don’t hold a candle to live production, the closest you can get is hydro storage, which not everyone has, and can’t realistically be built everywhere.

Look at the stats. The second largest battery storage in the US (and the world) is located near the Moss Landing Power Plant. It provides a capacity of 3000 MWh with 6000 MWh planned (Which would make it the largest). That sounds like a lot, but it’s located next to San Jose and San Fransisco, so lets pick just one of those counties to compare. The average energy usage in the county of San Clara, which contains San Jose (You might need to VPN from the US to see the source) is 17101 GWh per year, which is about 46.8 GWh per day, or 46800 MWh. So you’d need 8 more of those at 6000 MWh to even be able to store a day’s worth of electricity from that county alone, which has a population of about 2 million people. And that’s not even talking about all the realities that come with electricity like peak loads.

For reference, the largest hydro plant has a storage capacity of 40 GWh, 6.6x more (at 6000 MWh above).

Relative to how much space wind and solar use, nuclear is the clear winner. If a country doesn’t have massive amounts of empty area nuclear is unmissable. People also really hate seeing solar and wind farm. That’s not something I personally mind too much, but even in the best of countries people oppose renewables simply because it ruins their surroundings to them. Creating the infrastructure for such distributed energy networks to sustain large solar and wind farms is also quite hard and requires personnel that the entire world has shortages of, while a nuclear reactor is centralized and much easier to set up since it’s similar to current power plants. But a company that can build a nuclear plant isn’t going to be able to build a solar farm, or a wind farm, and in a similar way if every company that can make solar farms or wind farms is busy, their price will go up too. By balancing the load between nuclear, solar, and wind, we ensure the transition can happen as fast and affordable as possible.

There’s also the fact that it always works and can be scaled up or down on demand, and as such is the least polluting source (on the same level as renewables) that can reliably replace coal, natural gas, biomass, and any other always available source. You don’t want to fall back on those when the sun doesn’t shine or the wind doesn’t blow. If batteries were available to store that energy it’d be a different story. But unless you have large natural batteries like hydro plants with storage basins that you can pump water up to with excess electricity, it’s not sustainable. I’d wish it was, but it’s not. As it stands now, the world needs both renewables and nuclear to go fully neutral. Until something even better like nuclear fusion becomes viable.

Yes, but most people dont have that or take way too long than is worth in effort and (lack of) enjoyability for a simple meme. There already exist models to unblur entire images in seconds. AI should take the shitty work lol.

Let bro touch some feathers man 😭

Counter point: It’s from that one teacher who really gets teaching and it’s two hours of fun where you dont realize you’re learning