The No Subject Thread - The Bear and Badger

SpaceGazelle

Maybe we will experience a new Luddite era and people will fight back, but I doubt it.

bad_hair_day

There’s always hope, see the car thread.

poprock

SpaceGazelle wrote:

Supermarkets aside, what happens when AI can predict what you're going to do before you know yourself?

Welcome to the Department of Precrime. Our systems have determined that you are 96.4% likely to do a murder this afternoon. Your publicly-appointed defence lawyer has been notified and will call you within 60 minutes. Have a nice day.

monkey

SpaceGazelle wrote:

No but you never used to need a mobile either. It's going to happen.

They'll need to come up with some reason for people to buy them then. Something people want - not ads and spyware. Up to this point, all iterations of 'Computer Clamped to Face' have been niche (VR), non-starters (HoloLens) or disastrous (Google Glass).

poprock

Should probably shift this to the AI thread, but here’s the latest spinout from MIT, talking about literally doing this behaviour prediction stuff.

https://www.archetypeai.io/blog/introducing-archetype-ai---understand-the-real-world-in-real-time

LivDiv

We are definitely in an era of recognising a lot of this tech is shit, not good for us, mostly used to manipulate us, not necessary.

People are also much poorer than when mobile phones and smart phones blew up.

davyK

Predicting behaviour starts to be become troublesome when you consider applying for insurance.

SpaceGazelle

poprock wrote:

Should probably shift this to the AI thread, but here’s the latest spinout from MIT, talking about literally doing this behaviour prediction stuff. https://www.archetypeai.io/blog/introducing-archetype-ai---understand-the-real-world-in-real-time

Yeah. It's a long standing question but can you infer the rest of the universe and predict all its behaviour from say, a single apple? There are quantum limits because of probability (without cause) but on a macro scale it might not prove an issue. There are also limits on the sheer size of the universe, but the universe only seems to compute what's being observed anyway much like a videogame. It's efficient like that.

SpaceGazelle

davyK wrote:

Predicting behaviour starts to be become troublesome when you consider applying for insurance.

The main worry for me is manipulation to prevent future events. This could be useful for messages like "You're about to step in front of that car!", but more likely it's to dissuade you from buying a product you were going to buy in favour of something else. This manipulation will be non-obvious.

We do have laws against things like subliminal advertising but with AI it's too subtle to pin down and legislation is difficult when we don't exactly know what the model is doing, and what traits it's picking up and using for manipulation.

monkey

If a product can't reliably adhere to the law, the product is illegal.

monkey

How many threads has this argument spread to now? It won't stop until it's taken over the forum.

SpaceGazelle

monkey wrote:

If a product can't reliably adhere to the law, the product is illegal.

Proving that it's not adhering to the law is going to be the issue.

poprock

monkey wrote:

How many threads has this argument spread to now? It won't stop until it's taken over the forum.

We need AI forum mods to punt it back into ~~the long grass~~ the appropriate thread.

Wookienopants

Are we sure SG is in fact not an Ai?

monkey

SpaceGazelle wrote:

monkey wrote:

If a product can't reliably adhere to the law, the product is illegal.

Proving that it's not adhering to the law is going to be the issue.

That's easy. It would need to prove it's safe and legally compliant before it could be sold. If it can't, tough shit?

SpaceGazelle

One day this little forum will get a new AI member, and we won't know it's not a human. Maybe we'll never know.

SpaceGazelle

monkey wrote:

SpaceGazelle wrote:

monkey wrote:

If a product can't reliably adhere to the law, the product is illegal.

Proving that it's not adhering to the law is going to be the issue.

That's easy. It would need to prove it's safe and legally compliant before it could be sold. If it can't, tough shit?

Define "safe". We're heading towards a future where the bots rule and we won't know what the fuck they're actually doing.

SpaceGazelle

poprock

There’s still a big difference between all this tech generating info that people can base decisions on, and automating the actual decisions. Personally, that’s a line I don’t want crossed too quickly. And it’s where legislation will probably eventually sit.

monkey

SpaceGazelle wrote:

monkey wrote:

SpaceGazelle wrote:

monkey wrote:

If a product can't reliably adhere to the law, the product is illegal.

Proving that it's not adhering to the law is going to be the issue.

That's easy. It would need to prove it's safe and legally compliant before it could be sold. If it can't, tough shit?

Define "safe". We're heading towards a future where the bots rule and we won't know what the fuck they're actually doing.

There's laws that define what a safe product is.

Unless there's a good reason for people to buy AI, eye-scanning, mind-reading, hypno goggles, and wear them out to the shops, then they probably won't.

SpaceGazelle

Is Facebook safe?

poprock

I just saw an amazing thing. Looked like an ambulance, or a Police van, but the livery proclaimed it was actually the ‘Large whale disentanglement team’. What a job!

Wookienopants

SpaceGazelle wrote:

Is Facebook safe?

Nothing is 100% safe.
In the wrong hands a spoon can be lethal

Wookienopants

poprock wrote:

I just saw an amazing thing. Looked like an ambulance, or a Police van, but the livery proclaimed it was actually the ‘Large whale disentanglement team’. What a job!

Ok, topic change.
We need more on this

LivDiv

poprock wrote:

I just saw an amazing thing. Looked like an ambulance, or a Police van, but the livery proclaimed it was actually the ‘Large whale disentanglement team’. What a job!

The mum jokes write themselves

poprock

Wookienopants wrote:

Ok, topic change.
We need more on this

They’re part of British Divers Marine Life Rescue, according to Google. Which checks out - that was also written on the back of the van.

https://bdmlr.org.uk/about-bdmlr

From their timeline:
“2013: BDMLR sets up the UK’s first and only specialised Large Whale Disentanglement Team to rescue entangled whales at sea, with training provided by the Center for Coastal Studies and International Fund for Animal Welfare in the USA.”

Unlikely

What about the small whales. Do they just leave them to die? Bastards.

Pseudoscot

Wookienopants wrote:

SpaceGazelle wrote:

Is Facebook safe?

Nothing is 100% safe. In the wrong hands a spoon can be lethal

Its dull. It will hurt more.

RIP Alan Rickman

Brooks

I must say "computer says no" at least every other day about something computer. Sad.

Moot_Geeza

This must be the least Brooks post ever surely.

Howdy, Stranger!

Categories