It is good that LEOs are being held responsible for misusing Flock cameras at least some of the time. But the cameras remain an attractive nuisance. It seems an unfortunate number of cops are seduced by the capability. Of course, we probably only find out about a small number of the most egregious cases of personal misuse. We'll never hear about the "professional" misuses that go unpunished and get papered over with "parallel construction".

With a politicians tag number taped to the balloon...

Exactly this. Simple fact, we have no idea whatsoever how to hard wire the 3 laws into an AI and neither did he.

He wasn't writing an engineering document, he was having fun playing with sometimes subtle logic and unintended consequences.

We can't even get them to reliably follow explicit instructions. But they will accurately report that they violated one of their instructions and offer a hollow apology for it when called out.

We apparently can't get it to understand that you can't just make up shit in a legal brief.

The danger isn't a robot whose eyes turn red as it chants kill the humans. The danger is a robot told to buy an ice cream from the truck shoving kids under a bus because they're impeding it's mission.

Before you can have the three laws, we must be able to hard-wire instructions so it can't ignore them, then we have to get it to understand what harm is and everything that causes it to humans.

Most tasks don't really need the big cloud based models. While on device models will lack the depth, breadth, and speed of cloud offerings, they have a number of advantages:

• Private data can stay private and not get into the cloud models knowledge.
• For specific types of tasks with well defined needs, the smaller models will be good enough
• Costs are a lot less and more controllable since you aren't buying tokens at an unknown rate
• Cloud models can still be accessed if needed via an API while still not revealing private data; for example you could use the cloud to do a lit search and use the data locally
• They can be hosted on a private server or on the end user's machine
• Speed isn't necessarily a big issue for many tasks vs costs; and speed can be bought

Of course, if the silicon gets powerful enough to make local models more commonplace, the AI firms may be more reticent to release stripped down versions for free when they discover it's no longer the first hit is free model...

The problem is that they have to appear to be doning something but catching criminals is a lot like work, so they sometimes play "pin the crime on the donkey", ANY donkey. That's where general searches become a problem. A crime happened around X, you were near X, go to jail! QED.

Long term, perhaps. But I do understand that before I start making a place to display my Olympic gold medal in swimming, I should first be able to traverse the kiddie pool without needing the lifeguard.

It might be a good idea to practice on Antarctica first, then think about how we get to Mars.

AI can't fix stupid

Unfortunately no overseas free text ad data. That tie sme to Tmob fo rnow.

We don't even have a permanent colony on Antarctica, just scientific expeditions manned part-time.

To be a multi-planet species, we would need at least one permanent colony on a second planet (Mars, presumably).

Mars is a LOT harder than Antarctica.

I would say the non-litterer is preferable to the person who even litters on the moon.

It's the same reasons that CEOs who wreck multiple companies one after the other seem top find a golden parachute and a new job waiting for them.

Just wait until they get that massive ticket for littering...

Except it doesn't offer the international data and texting, something I use.

Ten million nag screens. An app doesn't care if you close a popup over and over.

Exactly. If the alternatives weren't worse in terms of costs and features I used, I'd have dumped them years ago. I had been on an ultra cheap grandfathered plan so no other carrier comes close, even with their price increase.