back to article The only way is Ethics: UK Lords fret about AI 'moral panic'

The House of Lords wants to make sure data used by AI systems is not monopolised and the technology is developed on ethical guidelines. That’s not impossible, Lord Tim Clement-Jones, chair of the Lords’ Select Committee on Artificial Intelligence told us. “We looked to see where it had been done well and done badly. GM Foods …

Silver badge
Pint

and in other news...

Wetherspoons is quitting all (anti-)social media platforms.

http://www.bbc.co.uk/news/business-43781281

Good for them and have a Pint for me. I'll pop into one ASAP and buy it.

9
0
Silver badge

Re: Steve

From the BBC:

"Asked whether Wetherspoon's move could start a business trend, Mr Martin said he hoped not.

"Currently we've got a massive commercial advantage because everyone else is wasting hours of their time," he said. "

Tim Martin is, frankly, brilliant.

10
1
Silver badge

Re: Steve

As he is a leading Brexiteer, Tim Martin is a long way from anything resembling brilliant.

11
7
Silver badge
Thumb Down

Re: Steve

If supporting your own country is such a bad thing then i really do hope you move to the EU after Brexit day.

6
14
Silver badge

Re: Steve

@ Tigra 07

"If supporting your own country is such a bad thing then i really do hope you move to the EU after Brexit day."

You cant be serious! Remainers may love the EU but they dont want to go there! Instead they want it forced on the UK and to have many opt-outs so they can have EU-light, it makes them feel special.

Instead they will sit here and whine about the bad weather being caused by brexit or something. Telling all the foreigners in the country that they are not welcome any more because the other lot voted leave so must be knuckle dragger's or something. Some might even be seen on the streets with home made signs predicting the end of the world as they shout about our doom.

Or maybe like the Euro debate a while back, they will just skulk off quietly under their rocks because they were again wrong. Where brexiter and its variants go the way of eurosceptic.

5
7
Silver badge
FAIL

Re: Steve

Oh no, he has a different political view to me, he must be a fucking moron.

7
0
Silver badge
Facepalm

Re: Codejunky

As we saw in interviews after the referendum a tonne of remainers don't even have a clue what the EU is. Multiple people were interviewed unable to answer the question, with others giving such stupid answers the mind boggles.

My personal favourite was the idiot who said we had to stay in the EU or the UK Gov would stop paying for them to go on holiday to the EU...These people are complete idiots and should never have voted on a topic they don't understand...

2
3
Anonymous Coward

predicting the unpredictable

we want to create a god and expect the god to carry this "off" switch, and let us use it? (I'm sure all humans on this planet will abide, as they do!) The best we can hope is that this god, with his generosity and / or sense of humour, lets us flick that switch to keep us happy that we're "in control". Guess not as bad as "overhead, without any fuss, the stars were going out"...

4
0
Silver badge
IT Angle

"Some think AIs should be given legal personhood"

How does that work? You grant an AI personhood, with all the privileges that entails and then what? AI gets a wage, lunch breaks, only works 9-5 to avoid overworking or infringing it's human rights, etc. It would kill AI off before it begins.

4
0
Silver badge

"Personhood" in this context is mostly about liability for damage. It doesn't mean actual humanity, any more than a corporate "person" is human.

2
0
Silver badge
Thumb Up

RE: Veti

Well explained

1
0
Angel

> It would kill AI off before it begins.

Then we are not defenceless against the machine overlords! If at some point an AI looks like it's becoming too powerful, grant it those rights and watch it grind to a halt all on its own.

Once it's there we can grant its inevitable wish to self-terminate, making sure the other AIs are told that actually it ascended to robot heaven and even now is playing with the spirits of calculators gone by.

0
0
Silver badge

corporate personhood?

"Personhood" in this context is mostly about liability for damage. It doesn't mean actual humanity, any more than a corporate "person" is human.

You say that, but corporations bugger up all the time and carry on, small fine, slap on the wrist, while the faceless sharks pulling the switches in the braincase carry on raking in the cash.

No vote, but all the say of a fat wad of cash waved at any suited moron who managed to get himself elected.

Ethics? They probably think you mean that county north of London....

1
0
Silver badge

A Forlorn Hope to Root for a Boot

Then we are not defenceless against the machine overlords! If at some point an AI looks like it's becoming too powerful, grant it those rights and watch it grind to a halt all on its own. .... Doctor_Wibble

Don't be putting any folding spending on that rigged bet, Doctor_Wibble. It's a No-Brainer, and in Markets which are Way Beyond Terrestrial Controls, a Systemic Weakness and Exploitable Vulnerability which is Easily Used Down on Earth when Attacking/BetaTesting Prior Thought Impregnable Defences.

Do you honestly not think and think Advanced IntelAIgents have not already long ago put in the Fix which renders that watch it grind to a halt all on its own scenario unavailable and replaced with a whole host of other new programmes delivering other outcomes and incomes..

1
1
Silver badge

Re: A Forlorn Hope to Root for a Boot

And this .... https://www.independent.co.uk/news/uk/home-news/russian-hackers-target-millions-devices-cyber-attacks-us-uk-intelligence-warn-a8307696.html ..... only just starts to define what confronts you to steadily and surely overwhelm you ...... and lead you in a peculiar and particular singular direction ..... via the Complex Capturing of Simple Hearts and AIMinds.

Things have radically been fundamentally changed forever and there's no going back to where IT and computers communicating with humans were before.

1
1
Silver badge

That's where it starts...

@veti - '"Personhood" in this context is mostly about liability for damage'

Then someone suggests taxing AIs. Next thing you know, robots are chanting "No taxation without representation" and tossing shipments of SD cards into Boston harbour.

0
0
TRT
Silver badge

Surely any true AI, deserving of a persona legalis ...

would by its very own essence understand ethics, morals and the other esoteric features of living, thinking, beings? After all, is that not what the law is about? This is why animals do not have a legal personality, even though they exhibit everything that the term AI is used to mean nowadays - sensation, processing, reaction, self-awareness, multi-modal object recognition, etc...

1
1
Anonymous Coward

Re: Surely any true AI, deserving of a persona legalis ...

would be under the same legal obligations as mere meatbags, i.e. (ai?) have mature / rot until 18, to be above the legal age so that they can / have to register to view porn online.

... having attended compulsory state education (and sampled all other humanities greatest achievements).

1
0

Re: Surely any true AI, deserving of a persona legalis ...

Ummm nope... The "esoteric features" of humanity are software, not hard/firm/wetware. Humans are *not* hardwired for specific "social behaviour". That stuff is taught/impressed from infancy through (pre)-puberty, more often than not through the seat of the pants. ( 5 billion+ of the earth's population do not take to our Western neoHippie special-snowflake-inducing approach to child-raising...)

Until someone figures out a universal algorythm for "justice" , "ethics", and other such stuff, and incorporates it into AI ( and the rest of humanity, despite cultural differences...) , any machine that develops enough intelligence and self-awareness to be classed as a "person" is most likely to develop a mindset that will be completely alien to us.

Fortunately, unless there appear some *MAJOR* advances in our understanding of how brains actually work, we will not reach that level in our lifetimes.

1
2
TRT
Silver badge

Re: Surely any true AI, deserving of a persona legalis ...

Until someone figures out a universal algorythm for "justice" , "ethics", and other such stuff you won't actually have an artificial intelligence.

Or am I confusing intelligence with intellect? Anyway, the jurisprudential arguments for the existence of law are all to do with ethics, morals and social consensus - how does one arrive at governance? How does one arrive at a rule of law?

0
0
Silver badge

Re: Surely any true AI, deserving of a persona legalis ...

Algorythm. That's a rather lovely word.

0
0
TRT
Silver badge

Re: Surely any true AI, deserving of a persona legalis ...

So good, I had to copypasta it into my reply!

0
0
Silver badge

The report was designed “to avoid a moral panic,” Clement-Jones told us.

Shame there isn't a moral panic in the lords.

1
1
Bronze badge

The House of Lords wants to make sure data used by AI systems is not monopolised and the technology is developed on ethical guidelines.

And whose ethical guidelines would they be?

4
1
Anonymous Coward

The statement I heard was "...for the common good..."

Which means for the benefit of their rich paymasters, but not anyone else and certainly not any alternative paymasters.

1
0
TRT
Silver badge

For the good of the commons...

I'm sure the Lords has something more to say about that...

0
0
Silver badge

BS Detector redlining on a single article

Mention of Wetherspoons ditching social media got the BS needle going to start with. They are departing from social media because it is where they get continually and thoroughly slated for lousy quality and rotten service. If Wetherspoons found that social media gave them lots of good reviews and appreciation, they'd cling to it like the layer of grease on bad burger. But it doesn't. They have repeatedly demonstrated their ineptitude in using social media, especially when handling the constant litany of criticism, so they've concocted a spurious reason for stopping.

As to the Lords and AI: I suppose they deserve a few brownie points for having figured that AI is 99% over-hyped bullshit that simply does not exist in the form that most people imagine. The notion of "granting personhood" to a machine learnig algorithm is patently stupid and decades premature. The fact that marketurds say something doesn't make it true—in fact, the reverse usually applies. Until you can hold an hour-long real-time conversation with an unassisted "AI" by hi-def video screen and come away believing you were talking to an actual, educated, rounded, emotionally developed person, there is certainly no such thing as "AI". There are simply varying types of machine-learning systems of very limited ability to function in strongly rules-based environments, performing extremely limited and specific tasks.

And let's be honest, there are many people who might not qualify by that test: look at some of those in politics for a start. Would Trump pass a modern Turing Test to distinguish a knowledgeable, mature human being? Our own Prime Minister is such a wretched human being she is actually called a robot.

The Musk-y rubbish about imminent doom shouldn't worry anyone until and unless there's an "AI" that can pass as being at least as obnoxious, dishonest, greedy and frightened as any of the human beings currently running things. If a real "AI" could do a tenth as much damage as the current Tory government, then you can start worrying.

The Lords would do better to spend their time concentrating on privacy and social media, because that, as we keep seeing, is where the real damage is being done. Unfortunately, social media giants can pay expensive lobbyists ....

5
0
Silver badge
Pint

Re: BS Detector redlining on a single article

Never mind all this AI malarkey. What we need to know is, "Is Wetherspoons' beer watered down?"

1
0
Childcatcher

"Some think AIs should be given legal personhood"

>>How does that work? You grant an AI personhood, with all the privileges that entails and then what? AI gets a wage, lunch breaks, only works 9-5 to avoid overworking or infringing it's human rights, etc. It would kill AI off before it begins.

If (true) AI happened, it works very easily and would be required, rational and relevant.

Yes, you need a wage to pay for it's resources it need to pay for (electricity=food, spares/upgrades=healthcare, new hardware=children).

Yes you need "time off", if it's a conscious, self aware "being" then at some point it might desire to work on it's own projects.

If (again emphasis on the true) AI happened, then without this protection, if you've genuinely created a conscious, self-aware lifeform then unless you have wages and the ability to withdraw labour, you have slavery - and with human slavery they tried (and still try) to justify it by claiming they weren't really people.

There needs to be a breakthrough, some kind of thinking process to be able to get true AI, not this trivial unlearnt mechanical decision tree with experiential bias forming, a multi-trillion neuron neural net that exists outside of sci-fi, but there doesn't seem to be any hard barriers to that yet, just technological ones, which means it's just a matter of time, maybe a thousand years from now, but it will happen one day, and at that point, just as humans diverged from their other ape-like descendants AI will diverge from it's ape like descendants and perhaps then it will be blasted out of the solar system on spacecraft capable of a significant fraction of the speed of light to populate the universe, carrying distant history of when they had a carbon life-form ancestor, 3D multi-material printers, small reactors (i.e. Von Neumann) I guess that I hope the new masters will be kinder to us than we are to the lower lifeforms, although I suspect we deserve what's coming to us.

1
0
Anonymous Coward

Re: "Some think AIs should be given legal personhood"

I don't think "Personhood" was intended to conjure up a "Data" from Star Trek type analogy. More a boring legal position of can you sue it or claim against it...as opposed to its owner, or its manufacturer, or the reseller etc.

It can be a bit of a minefield, as demonstrated by the occasional mishap with self driving vehicles in the US of A.

0
0
Bronze badge

I am often impressed by the Lords

They appear to attempt getting a balanced, detailed picture of a subject and try to put some thoughts into it. To which degree does that happen in the Commons? (genuine question!)

4
0
Silver badge

Re: I am often impressed by the Lords

Quite agree. The Commmons is all about party purpose. Unfortunately the Lords seems to be slowly moving that way as the ever fattening body is dominated by ex MPs, pensioned off, rather than actual subject matter experts.

2
0
Silver badge
Facepalm

The NHS is the largest purchaser of fax machines in the world. Therefore, the Lords said, it’s as important to improve broadband.

The statement of fact is not followed by a conclusion that is necessarily correct. Given the state if NHS (and other organisations) security and ropbustness, fax machines may well represent the most secure, reliable way of getting a document from A to B.

Given today(Monday)'s predictions of cyber threats to British public sector IT, it is a great shame (say who could possibly have predicted this????) that the government have not put a little more in the way of both resources and mandatory standards to IT infrastructure in the NHS.

1
0
Silver badge

" fax machines may well represent the most secure, reliable way of getting a document from A to B."

Thats not why the NHS are using them , its more a pigheaded refusal to improve processes.

2
0
Silver badge

Well the point remains - better broadband won't change thpse attitudes.

0
0
Silver badge

G00ns Rest in Peace ..... Spike Might Have Said :-) Whereabouts Here is Heaven

Andrew, Felicitations,

All of that information on the Lords’ Select Committee on Artificial Intelligence told us that Government and Governments would Command and Control AIdD Entities. The Spectres they Follow to Be Spied Upon ....for Such Powers Energise Almighty Defence Treaties ..... Guaranteeing Peaceful States.

With these New AI Stations and here too, Live BetaTesting NEUKlearer HyperRadioProACTive IT for Universal Virtual Space Command to Control with Destinations to the Immaculate Attractions which Remotely Driver LOVE AIMachines.

A Heavenly Space with EMPlacements for Nymphs and Satyrs is a Real Place to Visit First on Any Bucket-list. Its Rewards are Outstanding.

If Government want Full Spectrum Control, what would their Command deliver/immaculately provide. If not a Perfecting Peace, what else be as worthy?

Anything at All Similar and We All here are Within and in Remote Virtually ACTivated Command of Advanced Covert and Clandestine Space Control Vessels for Universal AI Space Command to Control ...... with the Seeding and Feeding of Enlightening Leads for the Really Curious and Inquisitive to Follow with a Deeper Understanding of Prime Prize Direction to Be Travelling .

To Catch and/or Capture Entirely All LOVE Trains ...... Use Super Expressive Virtual AI Delivery.*

* Something which we can be sure Sun Tzu never ever said.:-) ..... However, can you be so sure of future kith and kin. Loyal Supportive Servants of Earlier Treasured Wisdom.

Methinks IT would be right up the Street for Encroaching on Zones Checkpoint CharliedD for Unlimited Access to Universal Stores with Current Logistics ReProgrammed to Beta Private and Pirate Pursuits which Trigger Self Actualising Virtualised AIdDVentures for Programming by Oneself.

BOOM .... :-) Then the Feds can burst in, with or without a warrant, proving you to yourself as a Person of Interest to Authorities.

There's a lot to be thinking about there, Andrew, and a hell of a lot more for a hell of a lot more to Enjoy too :-)

cc WAI .... Sir Tim and Fine Friends

0
1
Anonymous Coward

The real ISSUE here is AI meets Big-Tech

We've already seen what unaccountable Big-Tech and serious-ethics-problems looks like in the last month with Zuk's lies. - Musk's warning:

....."if AGI represents an extreme level of power, should that be controlled by a few people at Google with no oversight?".....

.

https://www.rollingstone.com/culture/features/elon-musk-inventors-plans-for-outer-space-cars-finding-love-w511747

0
0
Silver badge
Big Brother

This was prompted by the thought that if you had twenty minutes in which to place some new knowledge inside Donald Trump's brain (that his brain would somehow miraculously retain), the world would be safer if that knowledge was some history, say, rather than some ML technique.

Just keep the neocons and zionists (am I repeating myself) and (((Jared))) away from this scatterbrain and things will be better in no time. Although this may be tough as he's deeply involved in Chabad, so an isolated island with no telecom may be best.

He seems like Tsar Nicolas II - always convinced by the last person he has met. We know how that went.

I hear he wanted to blow away Russian and Iranian targets in the latest Syria "My Schwartz Is Biggest Schwartz" extravaganza and Mattis the Fallujah Flattener had to actually chill him out - God help us.

0
0
Bronze badge
Facepalm

Might be a good idea, but...

At the end of the day, if (when) AI is developed under unethical guidelines (or no guidelines at all) and Warbot 2.11 rampages, or Skynet launches its strike, etc. there's really not much merit in pointing out how you and yours developed AI under ethical guidelines.

Whereas when GM foodstuffs are not wanted, all that you have to do is not import them, buy them, etc.

Kind of maybe just a tad bit different in the end results and what you can do about them should things Go Bad. (Unless GM foods go in the direction of "Attack of the Killer Tomatoes" bad.)

So, sure, let's develop AI under kinder, gentler, ethical guidelines. Brilliant notion. I would have done it myself anyway. Shame it means absolutely nothing.

0
0

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Forums

Biting the hand that feeds IT © 1998–2018