XorNot

It certainly matters as familiar sounding words to their stock holders to please not drop the valuation.

Because what they want them to think is "the AI factory has unique proprietary technology that cannot be replicated"

What they don't want them to think is "it's relatively easy once you know the basics to bootstrap to near SOTA and so the commercial case for selling inference has an extremely short profitability horizon with little if any brand loyalty or lock in".

They make and sell food in an age where literally everyone has a high quality camera in their pocket.

In their core business they're not going to show an actual picture of the actual product they intend to provide?

In an age where people routinely tweet/Instagram/message pictures of their served meals?

For food in addition to the false advertising it's just going to be the uncanny valley: there's thousands of years of evolutionary motivation to try and assess if food is safe. Any amount of "wrongness" will put people right off.

There's some hot take incoming about highly processed foods, but the reality is that's still designed and made by humans, for humans, and so passes all those initial hurdles like "is this poison or a trap?"

An image model doesn't even have that grounding.

The amount of time I spend per day waiting for my web browser to do about 10 redirects through Okta and touch a Yubikey is substantial.

This is marketing.

Frankly I'm inclined to say that it might also be faked: this drops just days after a new Chinese model does with the usual effect on OAIs projected stock price?

Trying to keep data on a USB drive hidden long term is a losing game is the thing.

If you're intercepted with anything of value what you want is to rapidly destroy it.

This was the big problem I found when I was trying to figure out a secure way use Yubikeys: they're both long term unreliable, and short term impossible to wipe or destroy fast enough.

Literally a printout of 1 time passwords has better security properties in an adversarial situation.

For a storage device what you want isn't a secure store, what you want is a pin pull or switch which wipes it instantly.

My personal hypothesis here is the Chinese government looked at the game and simply decided not to play:

An astute Chinese analyst could reasonably forecast that they had little chance of controlling the AI market due to sovereign trust issues, but would also note that AIs are just software.

When the dust settles the US still won't have factories, and the real value of AI models is still going to be embodying them and getting them to do real, consumer facing work.

Perhaps the most striking thing about the AI boom is how quickly the US abandoned the veneer of local manufacturing in favor of more expensive buildings producing nothing you couldn't make anywhere else on the planet...from imported parts.

I'll die on the hill that Helm charts are more of a problem because text/templating is a weird looking templating language then anything else.

No definitely weird.

Or just two to four robots working as a team.

God this whole comment section is weird.

It's like all the people who are so proud they don't use dishwashers.

I mean I like the much lower rates of dysentery and untreatable skin and dental conditions...

People have an extremely rosey view of the past shaped by the fact that period dramas are acted by modern Hollywood actors.

The real question is whether a hardware platform today would be good enough.

The model is software on a server or in the cloud, so the real question is are there any show stoppers from the hardware for going faster?

Okay but I cannot stress this enough: no one cares.

It's international politics. The rules are optional, and written on the back of whoever agrees to enforce them.

If you're going to run around declaring AI is a strategic advantage vital to national security, then guess what? Stealing it is a great idea. That you stole it is only a problem if it means you're not developing the ability to support that work locally as well, and China seems to be doing very well at building it's local talent and support network.

If you ever listen to Russian propaganda, there's a similar theme: every big idea, everything good, all of it was definitely first developed in Russia - only Russians could ever have thought of it. Of course, Russia isn't actually a world leader in any of those things, or able to execute on them.

Which is what America is sounding like more and more these days.

The hand wringing over whether internationally located AI labs are "stealing" output from American ones is the funniest thing in a while.

It's international politics with people talking about AI success as a matter of national strategic advantage and survival. So at best "this was built off our work" mostly tells you that apparently you've got months of advantage when a new model drops before it can be cloned. That's certainly some sort of advantage, sure hope it represents a consistent ability to stay ahead and causes people to redouble their efforts.

Or...of course none of these companies are worth what they say, but the advantage is also not really that great, and a whole lot of people are just really worried about their stock payouts.

See that detail of "near them"? Yeah it's a crucial one.

For some reason data center companies keep wanting to build data centers near existing utilities, near existing roads and transport infrastructure, and near existing electrical utilities.

Not in the middle of a desert. Or the antarctic. Or decommissioned oil fields.

After just not wearing a watch because I have a smart phone in my pocket all the time I'm out which has, amongst other functions, a lock on it and I can actually interact with one handed...I will continue simply not wearing an utterly redundant piece of technology.

Which does not change the fact that the pandemic was a peacetime crisis where it was possible but not practical to keep most systems running if needed.

Coronavirus wasn't bullseyeing vaccine shipments in the Pacific or taking down air freighters.

EDIT: I mean I don't know why you think this is a catch-22: countries pursue both capabilities, and the UK has a pharmaceutical industry and on shore manufacturing capabilities.

Whereas many industrialized nations are struggling to keep steel making capabilities on shore and running. So why is steel special? Because currently it's the one we're in danger of losing (and much harder to ship globally even if you have allies).

.. what were you doing in 2020?

Waiting several years for widespread vaccine availability, and practicing good hygiene and social distancing.

Viruses aren't so good at seizing and holding territory.

Vaccines aren't something you need urgently in a crisis.

A wartime effort has to keep supplies moving daily, or the front collapses.

Whereas the need for vaccines is heavily deferred - your population is already vaccinated in peacetime, and you are unlikely to need to make a novel vaccine over the course of a war, nor vaccinate new population during one either: that large vaccinated population providing herd immunity gives you a lot of runway for children with less access.

Because like, every basic war machine is made from it? Like how is this a question?

There's been two world wars and access to or stockpiling of steel has been a critical strategic factor in both.

Most consumable military assets are made from steel, for example and it underpins most machine tools and factory components as well.

Sending death threats has never been legal.

The statical lie with the UK numbers is the vast majority of cases are about the use of messaging services in domestic violence cases.

It has never been easy to create separate users on Linux, certainly not for tasks where you need to switch between contexts.

Docker was amongst the biggest steps forward on this in a long time.

Except it's fundamentally not equitable: regular people can't easily relocate, and use electricity to heat and cool their homes, cook food, and have lighting.

A data center can easily outbid them on the electricity market and drive prices through the roof for a few years, but also is free to simply turn off if it doesn't like the rates.

In theory yes. But in practice if an AI made sweeping changes to typography then what are the odds any effort at all went into the rest?

There was a virtual machine KVM escape found like 2 weeks ago.

Nothing is a security boundary anymore.

I'd like the Oauth authentication setting to work in the latest version. But that might just be me.

I'd also like it to be possible to enter a U2F pin number when using Oauth because then I could actually use it with my company Yubikey.

AIs aren't people. I can demand whatever I want from them.

The entire premise of building machines that can provide intelligence is that they can be arbitraryly rigorous and do far more work then is reasonable or practical to demand of a human.

Which is funny because a huge number of people post this in responses to online asynchronous messages, on public forums, when they could also do literally anything else - like close the browser tab.

LMGTFY for you was infuriating because it was 100% of the top Google responses to specific searches.