It's important to add in your simplified explanation that you'd include the frequency of the word pairs also. In ngrams, a primitive form of language prediction model, that would be all you need to store for a value of n=2. But for it to be useful, you'd want a value more like 4 or 5 at least. A value of n=3 would mean in addition to storing the frequency of word pairs, you'd store the frequency of word triplets (three words in sequence). n=4 is four words in sequence and so on. The size of the data store grows exponentially, but the realism of the output starts to get decent around 3 or 4 for a very simple algorithm that isn't even AI. Anyway. That's all.

Overfitting is an error, not a goal. And using lots and lots of different works is one common way to protect against it. It works similarly to how it would with a person - if you were to grow up only being able to ever read four poems, you'd easily recite them from memory, but if (assuming it were possible for a person) you read hundreds of poems and essays and books and articles every day, you wouldn't be able to remember any of them very well, but you'd have a very good idea of how language works and what range of written dialects make sense.

You might even have a basic idea of some styles, and might be able to create some original works in some of those styles, but at most a few words or phrases from some of the millions to billions of sources might come through.

Creating a 486 from scratch, rather than going through iterative phases of design and learning from the previous design and so on, is fairly amazing for an early stage AI technology. When this gets good it will most definitely change how chips are designed - probably as an assistant to design subcomponents rather than as something that does entire chips, at least at first. In time, this is a basic step on the way to self-reproducing machines that can improve themselves. Beware.

Not sure your prices are accurate, and you don't include food costs when comparing with the Marriott, but good post overall.

If they have an open source model, and released code / documentation to go with it for inference and training, they already did everything necessary for companies to make money off of it. They don't need to do anything else.

Why would they be considered securities if they donâ(TM)t grant you shares in anything, no control over any organization, no ownership of a piece of a company, no voting rights.. these smell like commodities to me.

Thank you for saving me the time. I can't believe you are the first one to mention this. Twitch is so stupid for doing that, at the same time they clearly admit the rules are bad for Twitch and it's users.

Excellent points.

So you are implying that workers never get treated like crap, and factories never have environmental problems in the U.S? Both things are catastrophic, nation-threatening problems here - the worker one at least being clearly potentially worse, since we have no healthcare system for people who aren't rich.

2.4% is not an insanely huge amount, and VR has a huge amount of potential that is strangely untapped. However, Meta/Facebook is trash, and always will be, and THEIR contribution to that 2.4% aside perhaps from some headset sales to help people get the hardware to access OTHER vr software that does not have the Meta/Facebook stain on it, will be negligible.

I would NEVER do any work inside something Facebook owns, I would never have a chat I didn't want published for all the world to see, I would never trust them or their platform to keep ANYTHING private. They literally have stated they don't believe in privacy. I don't even trust the E2E encryption on WhatsApp ever since they bought it. Maybe the lowercase meta verse will be 2.4% of the economy, but facebook's part in it (other than headsets) will be less than 0.000001%.

It's working correctly. This was most definitely planned and scheduled, intended to send everyone dumb enough to use edge's net traffic to microsoft until they got caught.

Itâ(TM)s vastly preferable for a citizen to do the shooting than for the police to do it. At least if a normal citizen does it, that citizen can be put on trial. If a police officer shoots you, youâ(TM)re just fucked, they will never be put on trial and they will never be convicted of anything, theyâ(TM)re basically immune to prosecution and can kill anyone they want for any reason. Police are effectively above the law and that is why they should never ever be doing the shooting.

This is not a troll. Itâ(TM)s just truth.

I mean, I would have paid at least 2 pounds.

There are no issues with storing unpressurized hydrogen, except how to get it out of the container if you literally mean 1 ATM. At 2 though, it will come out, youâ(TM)ll just get less than you would with a high pressure system. Youâ(TM)d need pretty incredible energy density for people to want to just not have much stored. Perhaps if you have something on the level of nuclear, but personally I donâ(TM)t see a gas mixture doing it.