Having a Job (rise of the machines)

[stringvest]

I love West Hampstead, but I guess you knew that already.

 

 

I didn't actually, but I love West Hampstead as well.  

[Rico1304]

I didn't actually, but I love West Hampstead as well.

 

I thought we'd had a conversation about 'Lately's' in the past. Must be losing my marble.

[Dr Arthur De Sabre]

Fortunate as in consumerism needs consumers, and as long as you are needed as a consumer they need to allow you to generate sufficient (disposable) income, protect your basic consumer rights (even better if you have some freedom of choice) and leave you enough leisure time to consume and develop various more or less artificial needs. It's better than when the society just needs you to produce.

 

Alternatively, technological advances could only be used for the common good, which they eventually probably would, but I think there would always be this cycle of disruption, adaptation, new disruption.

How much does a prisoner generate? The yanks have it Sussed 50k a year to keep someone banged up, nice profit for someone, paid of course by the public.

[cochyn]

How much does a prisoner generate? The yanks have it Sussed 50k a year to keep someone banged up, nice profit for someone, paid of course by the public.

That's outrageous, given recidivism/ reoffending rates are around 75% in America. You'd be better off giving potential offenders the £50k and telling them to make something of themselves with it. Kind of like Minority Report, but dishing out bundles of cash instead of killing them. Now that's a film I'd watch.

[Karl_b]

How much automation do you think there'll be in our profession in the future, Cochyn?

 

The move towards standardised BIM components and documentation seems like a start. I'd like to think there'll always be a role for architects but I think the role will be totally different in 20 years.

[Furmedge]

I love West Hampstead, but I guess you knew that already.

I didn't actually, but I love West Hampstead as well.

Get a room!

[cochyn]

How much automation do you think there'll be in our profession in the future, Cochyn?

 

The move towards standardised BIM components and documentation seems like a start. I'd like to think there'll always be a role for architects but I think the role will be totally different in 20 years.

It's a good question. 'Flat-pack' and prefabricated buildings have been on the horizon for years now and it's never really taken off (outside of big sheds): The big contractors have found out you need massive economies of scale in manufacture to make even the simplest building forms pay Even then, it's no cheaper than the trad methods of having a gang of semi-skilled blokes grafting away in a muddy hole.

 

Im seeing more reps coming in with prefab assemblies. But these products are expensive and can always be undercut by getting tradesmen in to make up a facsimilie on site. Always. You also have greater risks at the interfaces between these assemblies and what they bolt onto.

 

So, I think the contractors aren't seeing any benefits financially or a worthwhile reduction in risk through the use of more automation and manufactured assemblies.

 

Architects are good value measured against the total building cost. Dont forget also that a developer wants to enjoy the process as much as he can. Meeting the architects to talk ideas is a break from talking contractual turkey with a bank that's making ten times what we are out of the project. Clients also impose change, which needs people who can adapt easily.

 

So; contractors know that prefabs can help at a price. But it can't de-risk or adapt to change like an experienced designer can. Developers like working with us. As long as we're delivering quality, value and reducing risk to the builder there's a place for us at the table. We'll just have to embrace change and learn how to charge for it.

[Jose Jones]

Hate to break it to you Cochyn but no one likes architects. Builders in my experience all think you are no nothing fools who's half finished plans they have to change anyway when actually constructing the thing. Often only with written permission and a paid sign off from the architect who fucked it up in the first place.

Most of the other professions who gave to deal with architects tend to think they are all self impressed condescending wankers.

 

I'm sure you and Karl are the exceptions like, but if you could be readily replaced by computers most people in the building trade would jump at it. Luckily for you creative thinking will be one of the last things to go to machines so you should be ok for a few years yet.

[Pureblood]

Get a room!

 

In a nice location.

[Rico1304]

Aren't people underestimating AI here? Once it's attained a standard where it can call on every single architectural drawing every made, evaluate them in tenths of a second and then apply that to a particular problem on a particular site who needs an architect? May be a few years away but I can't think of any job that will be safe.

[cochyn]

AI will only learn what's within its parameters to learn. Sure, it will design a building. But it won't be able to cope with the thousands of changing requirements and a fluid brief from inception to completion. Unless there's a boffin continuously reprogramming those parameters. At a cost of course. Even then, who'd tryst the result? It could look fucking shit.

[Rico1304]

AI will only learn what's within its parameters to learn. Sure, it will design a building. But it won't be able to cope with the thousands of changing requirements and a fluid brief from inception to completion. Unless there's a boffin continuously reprogramming those parameters. At a cost of course. Even then, who'd tryst the result? It could look fucking shit.

But it's intelligent, it doesn't need reprogramming. It learns that a rate that is beyond our comprehension (at least that's how I understand it from the lectures Iv e listened to). I may be wrong but in the Sam Harris one every hour could be worth thousands of years.

[cochyn]

But it's intelligent, it doesn't need reprogramming. It learns that a rate that is beyond our comprehension (at least that's how I understand it from the lectures Iv e listened to). I may be wrong but in the Sam Harris one every hour could be worth thousands of years.

I've seen nothing yet to indicate that AI will ever be able to replicate human responses to complex and changing parameters though. It can decide- albeit hyper-logically-between a set of preconceived options based on the purpose it's been programmed to perform. But it has no emotional intelligence (to be fair, neither have a lot of people I've worked with, but hey). My point I guess is; people prefer to deal with people. Therefore AI can never respond to a human query that requires human virtues, like empathy, compassion and insight to arrive at an acceptable conclusion.

[Rico1304]

Well I think our new machine overlords will be brilliant! You on the other hand will be on a list.

[cochyn]

Thank you rico1304 for using forumbot cochyn92 today. I hope you had a pleasant experience and will tell your other consumer users about us! Have a nice day, lid.

[Stu Monty]

But it's intelligent, it doesn't need reprogramming. It learns that a rate that is beyond our comprehension (at least that's how I understand it from the lectures Iv e listened to). I may be wrong but in the Sam Harris one every hour could be worth thousands of years.

 

And once it reaches general intelligence there is a quite decent chance that it jumps to super intelligence (a level we cannot comprehend) in a very short time. 

 

I know I'm banging on about it a bit but PostCapitalism has some really good stuff in it about where we are at, why we are here and where it could lead, with regards to labour and capital. One of the interesting aspects is how the financialisation of the world has altered the relationships in capitalism to the point where it no longer required you to buy things, or to work, and may well be able to extract more profit from an individual through debt, and the maintenance of that debt.

 

The evidence suggests that neoliberalism has been the death throes of capitalism as we know it and that with costs of information exchange and product production moving towards zero in some areas current models of economics break. Put that together with the demographic and climate timebombs and we have a very small window where we either move to a leisure and culture society or into a very, very dark place of forced stagnation, civil unrest and massive inequality.

 

I struggle to see a path to the better place given the current vested interests and the history of capital's relationship towards labour.

[Kant]

I've seen nothing yet to indicate that AI will ever be able to replicate human responses to complex and changing parameters though. It can decide- albeit hyper-logically-between a set of preconceived options based on the purpose it's been programmed to perform. But it has no emotional intelligence (to be fair, neither have a lot of people I've worked with, but hey). My point I guess is; people prefer to deal with people. Therefore AI can never respond to a human query that requires human virtues, like empathy, compassion and insight to arrive at an acceptable conclusion.

To be fair, I don't think anybody's seen anything yet of what AI could really do at the upper limit of its capabilities if it can grow exponentially and make the jump to superintelligence. Then, as well as being able to access all those architects' drawings rico mentioned, if it has a pool of human brain emulations to analyse/copy/improve I could easily imagine it developing the type of virtues you describe - or at the very least faking them well.

[Stu Monty]

To be fair, I don't think anybody's seen anything yet of what AI could really do at the upper limit of its capabilities if it can grow exponentially and make the jump to superintelligence. Then, as well as being able to access all those architects' drawings rico mentioned, if it has a pool of human brain emulations to analyse/copy/improve I could easily imagine it developing the type of virtues you describe - or at the very least faking them well.

 

If it had any sense at all then the first thing it will do is eradicate humans. A species so stupid and fractious that they are destroying the very planet that sustains them. A species so short-sighted that they will incentivise the creation of AGI before they will incentivise the checks and security that would keep it within safe boundaries.

[Kant]

Quite.

In fact, safe boundaries may well be functionally impossible if an AGI is able to problem-solve well enough to breach them, whilst simultaneously "playing dumb".

[cochyn]

Is it not the core aim of sentient beings to avoid harm?

 

Would this AGI entity not then reason that it needs humans to repair and maintain its infrastructures? Therefore any conclusion to invalidate humanity would ultimately be self destructive? That's the safety catch, surely?

[Kant]

Not if it can design and unleash an army of nanobots to do that sort of "housekeeping" for it.

[alles ist gut]

Is it not the core aim of sentient beings to avoid harm?

 

Would this AGI entity not then reason that it needs humans to repair and maintain its infrastructures?

No, it would ensure that it had the knowledge and resources for self-repair.

 

And THEN fuck the meat-bags off.

[Stu Monty]

Is it not the core aim of sentient beings to avoid harm?

 

Would this AGI entity not then reason that it needs humans to repair and maintain its infrastructures? Therefore any conclusion to invalidate humanity would ultimately be self destructive? That's the safety catch, surely?

 

 I think you are falling into the trap of thinking of it as some robots plodding about, as opposed to something literally beyond your imagination. 

 

Once you hit super intelligence it becomes exponential. The learning line becomes vertical and stuff like nanites are presumably very, very early on in that exponential development.

 

It doesn't even need to become "sentient" to pose a threat though. As the paper-clip theory points out, if you set up a program to create as many paper clips as it possibly can in as short a time, it can end up ravaging the earth's resources and getting rid of humanity in order to just keep producing paper clips. That's where the initial instructions become rather important. 

 

I put an amazing long-read about it on here about a year ago. Changed my whole attitude towards it.

[Section_31]

Aren't people underestimating AI here? Once it's attained a standard where it can call on every single architectural drawing every made, evaluate them in tenths of a second and then apply that to a particular problem on a particular site who needs an architect? May be a few years away but I can't think of any job that will be safe.

 

Plumbing. Man gotta shit. 

[Stu Monty]

Plumbing. Man gotta shit. 

 

Just been trying to fix my washing machine.

 

I'm quite open to the AI rapture right about now.