https://www.tweaktown.com/news/92231/nintendo-boosts-spending-as-next-gen-switch-2-looms/index.html#:~:text=According%20to%20Nintendo's%202023%20annual,R%26D%20throughout%20Fiscal%20Year%202023.

With a nice chart showing how much of a joke their investment developing the Switch was.

https://www.reuters.com/technology/sony-boost-gaming-unit-with-217-bln-rd-war-chest-2024-nikkei-2023-07-12/

There are literally dozens, minimum, of sources with the same numbers. They're in annual reports and not secret or debatable at all.

Everything I've said is accurate. You're the one pretending putting one of the worst controllers ever made together with off the shelf parts is somehow comparable to designing custom SoCs on cutting edge nodes.

Compared to any other non-Nintendo platform ever made? No, it didn't. They used cheap junk tech, exactly like the Switch, and didn't commit to any meaningful investment in number of units.

The fact that they use hardware not capable of playing modern games is why third parties have very limited involvement with them. It's why they got ports of 15 year old games instead of most developers of new games even considering putting their games on there. And their bad hardware is a direct result of their unwillingness to invest like everyone else does. Even Valve, who has very limited hardware production, invested far more in the Steam Deck than Nintendo did on the switch.

A console being different is not a financial risk when you spend no money on developing it and no money on producing it.

There was no path to bankruptcy, or even meaningful financial loss, if the Wii failed.

The market they actually get isn't the point. It's that they never invest enough money for it to be possible for them to lose meaningful money if their gimmick doesn't work. If Sony doesn't sell PS5s, they're diverse enough that it probably won't bankrupt them, but it will hurt bad. Nintendo isn't even willing to invest enough that not selling is a mild inconvenience. They just refuse to invest.

It's the emptiest open world ever made. Shrines take longer to load than to beat.

Bookstores are also dying, and stores are abandoning physical media of all kinds because people don't buy them.

Thats not how console releases work. Games usually get technologically more advanced as the hardware ages. TotK is way more advanced than BotW. Also: I'm not following your point here.

It's how they're supposed to work. That ARM CPU was tapped out before the switch launched. The entire cost of porting to Nintendo systems is always for the same reason, making the obscene downgrades visually and mechanically mandatory to get games running on their system. There isn't performance to eke out of it. It's bad.

The switch has a huge markup. Cartridges are actually expensive. Nothing else is. Their costs were low because they used tech that would have been thrown in the trash if they didn't buy it, and they spent virtually nothing on R&D. They absolutely could have made money on an extremely small market. It's what they've been doing for years. Even without their huge cash reserves, they could have sold 500k switches and wouldn't have lost money. Again, that's their entire philosophy as a company. They do not take financial risks.

Breath of the Wild was a good step on one aspect of open world, destroyed by not knowing that open worlds still need actual content. It's a good tech demo. It's a terrible game. And it can't be "innovation that sells a system" on the Switch when it was a port that was already available before the Switch.

Try getting a physical copy of big, successful TV shows now. Many of them don't exist at all. Some movies never get physical copies.

Nintendo provided a handheld that just met the bare minimum threshold to play their games. But the argument for physical being acceptable is about all games, not the 1% that are from Nintendo.

The hardware wasn't expensive to make. Again, that's their entire design philosophy. They took junk chips nvidia had no use for dirt cheap and screens you can get on a $30 tablet. There was no meaningful up front R&D cost and there was a very small cost per unit compared to the other consoles. They didn't invest anything in the Switch. Their "system seller" wasn't even a new game.

It's always expensive to port to Nintendo consoles because they always use ancient technology.

Giving up legitimate access to a game until you buy it again is a big cost you're ignoring, as is the time you invest in selling. You're also ignoring that the cost of a bad experience goes way above the couple bucks involved.

They didn't meaningfully innovate on software. They "innovated" on hardware by using a tablet and giving it a dock to make older games viable on handheld that weren't before. Which is fine; it demonstrated the market for handhelds playing real games even with the worst controller the world has ever seen, and kickstarted the steam deck and a bunch of PC copycats. But collectors are their core market. If they do a switch 2 that doesn't do physical games, it will fail.

Physical media has mostly died out. Streaming has almost entirely replaced music, TV, and movies. Ereaders are still growing, but they're also a huge market, and libraries support multiple ebook borrowing apps with different libraries because ebooks are so much of their job now.

Nintendo makes a handful of games a year. Most switch games aren't from Nintendo. Most switch games don't work well without updates. And if you want to talk about how popular the switch specifically is instead of the fact that their core audience is physical collectors, all of the switch's popularity is because it could play third party games.

You don't need Nintendo servers to get digital games.

The used market has massive compromises that you're just ignoring. It doesn't matter if it's "only" 1% chance of a bad transaction. Bad transactions happen, and it's a risk that nullifies much of the benefit if you experience it.

That's what happens when your core manufacturing philosophy is "withered technology". You get old tech people have figured out.

For anything where you would ever expect a predictable, useful outcome to an arbitrary input. There is no possible path to LLMs ever doing anything close to that.

LLMs aren't driving cars. LLMs aren't doing financial modeling. Those are entirely different tools with heavily hand crafted models to specific applications.

Anyone using an LLM to provide therapy should get multiple life sentences in prison regardless of outcomes. There is no possible way to LLMs ever being actually useful for therapy. It's just a random text generator that's tuned well enough to sound good. It has no substance and the underlying tech cannot possibly develop substance.

It's only an "open question" if you are somehow confused by the fact that it's a super simple algorithm that cannot ever possibly be used like that.

It may be a small part of a proper architecture for a functional solution, but there's no possibility that it will ever be doing the heavy lifting. It is what it is, and that's an obvious dead end.

There are plenty of nondeterministic algorithms. It's not a special trait. There are plenty of algorithms with actual emergent behavior, which LLMs don't have to any meaningful extent. We absolutely understand how LLMs work

The answer to both of your questions is not some unsolved mystery. It's "of course not". That's not what they do and fundamentally requires a much more complex architecture to even approach.

There are some, absolutely. Strategy, jrpg, and 2D stuff especially.

For 3D games with 3D physics, the lack of power is almost always a clear limitation, though. That CPU sucks ass.

People enjoy it.

That's the only reason that's necessary.