Having to deal with a drug dealer that wants to also sell you actually addictive drugs is the gateway.

Marijuana is addictive though. Maybe not as addictive as some other things, but pretending it’s completely non addictive is disingenuous and misleading. It’s more addictive than say LSD or psilocybin for example.

That isn’t to say it should remain illegal though. Legalisation has positive benefits even for harder, more addictive substances than marijuana. See the history of alcohol prohibition for example, or the disaster that is the war on drugs.

If all it’s running is a hot tub that sounds reasonable. This bitcoin miner uses over 3kW: https://www.aozhiminer.com/showroom/high-profit-110th-95th-bitmain-antminer-s19-pro-s19-bitcoin-miner-with-power-supply.html

Pretty much all artificial neural nets I have seen don’t do all or nothing activation. They all seem to have activation states encoded as some kind of binary number. I think this is to mimic the effects of variable firing rates.

The idea of a neural network doing stuff in the background is interesting though.

A large data centre can use over 100 MW at the high end. Certainly enough to power a swimming pool or three. In fact swimming pools are normally measured in kW not MW.

What sort of differences are we looking at exactly?

I think he means the cat

Depending on what battery protection modes are in play, many have smart charging or other features designed to prolong life. Also a fair few batteries come out with greater than design capacity from the factory. It’s called a design capacity and not an absolute capacity for a reason. A phone battery that left the factory at 110% could conceivably still be at or above 100%.

Fyi it’s not overnight charging that’s the issue either, it’s charging to 100%. What one device consider 100% varies and devices will essentially lie to you about it. 4.2V is normally considered 100% full for Lithium Cobalt Oxide batteries yet some devices push higher than this while others skirt under to pad capacity and cycle life respectively. It’s about tradeoffs.

Man I hope so but you really never know these days

Plenty of alternatives to Windows lol. Try a different industry.

It’s not just google who have AI stuff built into their phones. All recent SoCs I have seen have had NPUs going back the last couple generations. A lot of older or cheap phones won’t have one, but the new devices will.

I don’t see the problem with using the phones normal GPU. This shouldn’t be more insecure than making a call currently is. I am pretty sure android phones don’t have a secure enclave just for making calls as you can give different apps access to calling features, and most calls I make are through third party apps anyway, not via POTS. That being said android is pretty secure anyway provided you don’t give permissions to the wrong app. It’s more secure than your average Linux system, as each app has its own user and is only allowed to access things it has explicit permissions to access. Secure enclaves aren’t all that in my opinion.

And let’s not forget that if the phone can listen to your conversation to detect malicious intent, any country can legally compel Google to provide them with the data by claiming it is part of a law-enforcement investigation.

The point of doing it locally is the audio never gets sent to google directly. That being said they could definitely do some dodgy things by training the ML model to search for words like abortion, drugs, transgender, etc depending on what the laws are in the country the phone is being used in.

Someone confusing load-store with RISC again.

So you do include ephemeral state that’s a copy of database data? If we were including that then every non-static website has plenty of state, but so does every web server. Whatever definition you are using must be quite odd.

I don’t write web applications for a living and I especially don’t write front ends. I do have to ask though:

What information are you actually keeping in the front end or web server? Surely you don’t need any ephemeral state that isn’t already stored in the browser and/or for you like the URL or form details. Only thing I can think of is the session ID, and that’s normally a server side thing.

I mean I’ve written web sites where there is no JavaScript at all, and the server is stateless or close to it. It’s not a difficult thing to do even. All the actual information is in the database, the web server fetches it, embedds it into a HTML template, and sends it to the client. Client doesn’t store anything and neither does the server. Unless I really don’t understand what you mean by state. You might keep some of your server fetches data from another server using REST or SOAP but that’s only used once as well.

The whole conversation was about backend being similar because you can write a stateless server. Have you forgotten? The issue here is a backend isn’t just one service, you can write a stateless service but you are in fact just moving the statefulness to the database server. The whole backend isn’t simpler than the front-end for that reason. It might be simpler for other reasons, though many popular websites need complex backends.

I am not arguing that a stateless service isn’t a useful concept. I get why people might want that. That’s not what this conversation is about. It’s about the backend vs frontend. Backend to me includes databases and other support services.

You still have to consider ACID vs BASE when choosing a database software/provider. It comes from CAP theorem.

If that’s how people want to talk about it they can, but you can’t eliminate statefulness from the whole stack.

Yes that’s a stateless service but not a stateless backend. A backend to me is everything that doesn’t run on the client, including the database. Databases are not stateless, even distributed databases are not stateless. You can’t just spin up more databases without thinking about replication and consistency.

Sure

The only nuances here seem to be: a) very simple websites need little state (but still aren’t stateless) and b) that you can move the state around to make something look stateless within a narrow view.

16 GB should be the minimum for a new PC these days. The fact they are selling less takes the piss.

It didn’t realize I was replying to someone else. The person I thought I was replying to was on a 4790K.

This change isn’t a problem at all, just like the Nazis interior decor wasn’t a problem. AI needing certain instructions to run well isn’t unreasonable at all. They are using these instructions and resources to provide a service. This isn’t them wasting resources unnecessarily, presumably they are only used if you engage those services. Don’t get me wrong Microsoft does waste resources but this isn’t an example of that afaik.

What you and others are doing is a motte and bailey argument. First you say these AI requirements are unreasonable, then I say actually they aren’t unreasonable at all and is well exceeded by the actual Windows 11 requirements or most machines made in the last 10 years, then you counter that the Windows 11 is slow and has unreasonable requirements. Do you see how the first and second points are unrelated? The general Windows 11 requirements are way more strict than requiring SSE 4.2. any PC back to like 1st or 2nd gen will have the needed instructions. Maybe the 16 GB requirement is a bit more than some people have, but it’s not a large amount either and you can just keep using it without the AI features.