You're probably looking at a choice here because most systems will freak out from the noisy brute force attacks and start alerting... cue the Gary Oldman meme... EVERYONE!!!! within minutes.

One of the first things any attempt to restore democracy must do is use antitrust to destroy the near monopolies.

That's also the only way to save Capitalism from Monopolists.

The fuzziness is intentional because they don't want to be on the hook for accidents if they can possibly avoid liability.

My favorite way to pick on Tesla owners is to tell them that AutoPilot, Full Self Driving, and Full Self-Driving Supervised to not allow for ANY truly hands / attention free operation because you need to be able to take over driving in an instant. You are ALWAYS liable if the car gets into an accident.

On the flip side Mercedes Drive Pilot, under specific circumstances, actually allows you to go completely hands / attention free, and will give you a ten second warning to take over if necessary. Mercedes actually takes legal responsibility and indemnifies the driver while Drive Pilot is engaged.

Looks like Tesla doesn't believe AP or FSD is safe, almost a decade after Elon declared it would be able to drive you from West Coast to East Coast without touching the wheel.

As a pilot I'm all for automation, but we need to have the ability to disable automation completely in moments. I'm not talking about "press the button to turn it off" here. The first thing you learn about the autopilot during training is where the fuse / circuit breaker is located. Why? "So if you ever even
THINK you are fighting with the autopilot that's not showing an error you can immediately disable all functions in one simple action, lobotomizing the system.

We need the same ability so people driving a car can immediately disable any automation.

Sorry man but this is like Network Neutrality: I paid you for a product with features. Now if you want to claim you can't do it for the price offered then raise the price, stop selling the product, and/or file for bankruptcy.

This is how disastrous click-through EULA's get false justification and it's pretty much a "We are 'selling' this product to you but we can do whatever we want and if you don't like it, tough. All of our competitors are doing the exact same thing." Create hundreds of pages of legalese that no reasonable lay person can even be expected to understand, with clauses buried in clauses, and worse. So what's worse? Even when they scream and demand their contracts be honored they often fight tooth and nail against following far simpler open source requirements.

We need to stop normalizing this bull. This is dumb as bricks and should be ripped out of every single car that was subject to it... along with the unwanted telematics being reported to the mothership... or if we want to be generous having that information completely illegal to sell and treated as PII. That means just possessing things like location data become a liability instead of a possible profit center.

This is the same as the Mythos found tons of bugs problem. Sure Mythos found bugs, after being directed by security researchers. Oh and a lot of the bugs, yes the super dangerous bugs as well, were pretty much unusable.

The problem for OpenAI and Anthropic is they desperately need to keep the hype train on the rails until IPO at a time when more people are going, "What is this 'AI' stuff actually capable of doing without supervision?" Then on top of that some people are also beginning to ask, "Is there a good return on the cost of deployment?"

If you need to pay humans to babysit the model, and double check everything it does, and they get worse at double checking the longer they use the model, then is there an actual ROI?

Yes I'm just noting for people that many in the "Cult of LLM" doesn't realize the "church elders" (Sam / Dario) are running a game right now.

For those who don't understand, think of an MLM because that's the game right now. Get that initial group (VC's) their payout FAST before the collapse. Elon beat them to the punch, and I bet Sammy and Dario are P-I-S-S-E-D right now.

This would be different if it were truly transformative... but it isn't so we are just watching them speed run to IPO.

Nope and it's not AI. Those are expert systems and the chatbots are Lizabots making Barnum statements.

The biggest problem OpenAI and Anthropic have is they don't have AI, they have a crappy version of the Library Computer Access Retrieval System (LCARS) or a super Wiki. Now that could be useful, if it always knew that 1 + 1 = 2 but due to the fact it has terrible data inputs and "hallucinates", the accurate statement would be pulls incorrect data or calculated incorrectly, constantly they've done something I didn't know what possible. What did they do? Make a less factual version of Wikipedia.

While some people try to quantify LCARS as AI, it doesn't do anything unless it's told. There is no intelligence. The thing is the sum of all knowledge in the Federation databases. Now if Sam and Dario did that, with all the modeling and so on, that LCARS could do it would be a technologically impressive feat. That still wouldn't be AI, and it wouldn't be worth the TRILLION PLUS dollars of CapEx blown so far. The fact they even say they CAN'T fix the major flaw is telling that inference models are a mathematical dead end.

Dressing up 1960's technology with a fresh coat of paint and faster hardware is not AI.

Who's going to pay for them if the hyperscalers if the data centers are completed? OpenAI and Anthropic can't come close to generating the money needed to pay for all that compute. There is nobody else that can even come close if those two can't afford the bill. This is also completely ignoring the fact that just getting parts for things like, say, a natural gas power plant are now on a 5 to 6 year backlog.

Statements like these are only performative to keep the AI hype train going long enough that OpenAI and Anthropic can try to get to IPO before the bubble pops. Why? Because Sam and Dario don't want to break the one rul in Silicon Valley: Thou shalt get to IPO so VC's can get their exit liquidity, no matter how terrible the product / idea.

They know that if they fail, due to the vast amount of money involved, they will be forever excommunicated from the Church of Capital.

This.

My family has had a lot of people in the trades from carpenters to masons to electricians to machinists. The work will break you. I did concrete work as a kid and learned really quickly that I didn't want to spend more than 10 years (much less 40) breaking myself to pieces. Even the years I did concrete initially to help pay for college caused damage. My BiL is a carpenter who spent 50 years being a carpenter... and he's got so many physical ills I can't imagine his life.

Plus there's another brutal reality here that they're ignoring: Getting someone from apprentice to competent journeyman takes thousands of hours of actual work - often being watched by someone who is two grades above - or a few YEARS. My cousin spent his first year as an electrician's apprentice doing every possible stupid / annoying thing so that the guys who were close to journeyman wouldn't have to deal with it and could actually work. Think opening all of the cable, prepping all the fixtures, taking all of the boxes out of the boxes, cleaning tools, and every other possible annoying / stupid thing. How would you like to climb up and down ladders to install a few hundred light bulbs in an office?

The second year he was doing actual work but most of it was the low risk items - example being standard wall outlets. Ever wonder what it's like to crawl around a giant office building on your knees wiring a hundred plus wall outlets per floor? How about climbing up and down a ladder to install all the fluorescent light fixtures / ballasts?

I don't think these companies thought their cunning plan all the way through... because apprentice to journeyman can easily take 3 years.

I can say one thing: They made it tolerable.

Keep in mind there is no actual crushing economy on that flight - the seats are all premium economy seats or business. This looks to have roughly the same number of seats. Hopefully that is because Qantas have also have gotten rid of cramped economy.

However let's not pretend that, even in business lie-flat seats, that this is comfortable. I've flown in both on the Singapore Airlines route and business is only marginally better than premium economy. Most of that isn't the seat, it's the space to stand and walk around or the meals. Yes it is easier to sleep flat, but on an 18+ hour flight you sleep a couple times... and that seat is still designed to be a seat... not a bed... you will get up and move around.

Best of luck for all that fly that route - I hope it's tolerable. That's about the best you can expect out of being cooped up that long in a plane.

There is no functionally useful "AI" that is truly cheaper.

They have to get exit liquidity somehow. They can't get it if people start actually asking hard questions. They can't make a chatbot look super smart if people start asking why it can't do what they said it would years ago. They can't be allowed to ask why people still have programming jobs even though so-called AI is supposed to have wiped out that job sector already. They can't be allowed to ask why people who use so-called AI don't work any faster, and cannot troubleshoot nearly as well. They can't be allowed to ask "How is this profitable?"

At least not for a couple more cycles. The suppliers (Nvidia, Samsung, SK Hynix, et al.) need to make good on a couple production commitments to get bonuses. Sam and Dario desperately need IPO's to make sure the VC's get exit liquidity or they will have committed the only sin in Silicon Valley. They will have broken the commandment:

"Thou shalt get to IPO and ensure the glorious Venture Capitalists get paid, or thou will be excommunicated from the Church of Capital."

Not really they are looking for money to keep the train running until they get exit liquidity. Nvidia is looking to keep the hype train running because they get money in the long run if they can keep the lemmings running in the same direction.

Although the real dirty words here are "circular financing" because Nvidia is saying they'll give up to $250B to OpenAI to build datacenters if nobody else will pony up. That's on top of their OTHER so-called commitments folks.

Where's the money coming from for these deals? Are these deals all vapor to keep the public believing in what is falsely labeled as AI?

I think the harder thing for more people is the cost. When I was a little kid our babysitter could take us to see Star Wars (or other movies) at the downtown Marcus discount theater for $0.25 a kid. Even making minimum wage she could take a car full of kids (us and our friends) for a buck. Then she could get soda and popcorn for those kids for less than $5.

As a teenager I could take a girl out on a date for less than $6 for a Friday night movie, then add another $8 for soda, popcorn.

Today my wife and I spend $40 on tickets and another $30 on popcorn and soda.

Don't get me wrong, Marcus does a pretty good job at making it a comfy and technologically impressive experience. The seats are nice, far nicer than the plastic "school theater" seats of my childhood. That makes a long movie far more tolerable than in the 1980's or 1990's or 2000's.

The concessions haven't changed, the popcorn hasn't really changed if you go to the right Marcus theater where they still pop everything on-site. That's where the cost bites the most... but it's also about the only place the chains can make any money.

Just saying... The chains (Marcus at least) have tried to adapt to the times. For most people, especially parents and kids, it's the cost that's more painful.