This topic contains 28 replies, has 16 voices, and was last updated by  don 2 weeks, 4 days ago.

Viewing 29 posts - 1 through 29 (of 29 total)
  • Author
    Posts
  • #3548

    Niles
    Participant

    If think AI can be a good thing but NOT in the hands of the tech elite (plutocrats) in charge at the moment. I think that it is just too naive to assume that the silicon valley elite will want to share. The big tech company’s don’t even want to pay their fair share of taxes most of the time and have had a devastating effect on small businesses, create of exploitative employment systems such as uber… etc. Their current attitude seems is one of a greedy toddler who wants everything and anything but will throw his toys out the pram as soon as he doesn’t get it (Donald Trump). This means it either their way or the highway and we will ALL have to beg for scraps from them.

    What do you think is AI going to be a good thing or a bad thing?

    #3549

    Adam
    Participant

    I agree silicon valley nerds of EVIL! Just check out this story of a silicon tech boss complaining about fruit sellers on the streets and wanting to get rid of them https://www.theguardian.com/technology/2016/may/02/mark-woodward-facebook-fruit-vendor-comments

    #3551

    Muhammad
    Participant

    Um.

    This “elite” that you talk of want to develop enabling technology that they can then sell to the masses in great numbers in order that their lives improve and they make loads of money.

    You should read some biographies of the likes of Steve Wozniak, Steve Jobs, Zukenberg, these sorts of CEOs before labeling them as evil manipulators of the masses.

    #3552

    Niles
    Participant

    @muhammad Hardly they simply wish to control and be at the top.. if they really wanted to make a world a better place they would start by paying taxes as that is one way to help society… but because this isn’t on their terms and what they want to do they dont. Although Steve Wozniak, Steve Jobs, and even Zukerberg aren’t the worst of the silicon valley bunch I have to admit. I can think of a lot better things to do then read a biography of someone talking about all the good deeds that they have done like a robber baron.

    • This reply was modified 3 weeks, 1 day ago by  Niles.
    #3554

    jack
    Participant

    @muhammad you are a typical liberal douche bag.. you don’t care about the effect that these big company’s are having on peoples lives right now and only care about your own cushy life.

    These tech company’s are crippling small businesses mercilessly.. practically using slave labor avoiding tax and paying our governments off. BUT WHO CARES AS LONG AS YOU GET THE LATEST GOOGLE HOME OR IPHONE!

    #3555

    troll
    Participant

    @niles Do you think if all the wealthy US tech firms paid a bucket load more in tax, that it would be spent on the welfare of the US people?

    I don’t.

    #3556

    Niles
    Participant

    @troll I agree that the US are very stingy when it comes to helping people in need.. but that is their problem. These big tech companies avoid tax not just in the US but across the whole world in countries strapped for cash. For example a couple billion that google avoids paying because of a deal they struck with George Osbourne could have gone a long way in helping people in the UK when government budget cuts targeting the most vulnerable in society.

    Putting blind faith in a “few” big tech companies is not good idea, the bosses that run the company’s are not saints for example if you look at people like Eric Schmidt (he is a sex pest and Nazis). I wish there was MORE competition in the tech world so that the wealth is not just concentrated in California.. and that will mean even more powerful growth that benefits more people. BUT these tech company’s like google don’t want this and game the system by paying governments and making deals to stifle competition so they remain in control. Bad not just for people but also innovation.

    • This reply was modified 3 weeks, 1 day ago by  Niles.
    #3558

    Peter
    Participant

    Knowing senior managers in Apple and having a son who studied engineering at Berkeley I know that “some” of these people are simply geeks who love what they do and like the idea of making loads of money while changing the world a bit. In my experience anyway.

    All companies, all companies, pay as little tax as possible and use their financial power to keep ahead of the competition. That doesn’t make them evil, it just makes them normal, a bit like big dumb animals. It is up to governments to regulate their behavior through anti trust legislation etc. I would love to see another silicon valley in the UK but we don’t have the critical mass of technology companies nor enough high tech investors willing to take a punt on 20 or 30 start ups in the hope that one will succeed and give them a good return. That’s why I support the Labour party’s policy of starting an investment bank to support young UK businesses.

    Maybe you know something that I don’t though.

    • This reply was modified 3 weeks, 1 day ago by  Peter.
    #3560

    Muhammad
    Participant

    It’s a very US type of attitude and one that I admire. Someone decides they’re going to change the world and then figures out how they’re going to do it who cares if they step on a few people to get on top. The thing is, there are enough people in the US with money who are willing to fund these ambitious projects.

    Ellon Musk is another worthy of mention.

    I guess another misconception is that things like autonomous vehicles, expert systems and the internet of things are all being driven by Californian companies when of course they’re not. Who ends up as the winners in autonomys vehicles is anyone’s guess. It could be Ford or BMW, Bosch or Nvidia or Alphabet..or most probably some kind of alliances between different companies pooling their knowledge.

    The future is always over sold and under imagined!

    #3561

    erick
    Participant

    I can’t see how it can possibly end well.

    #3562

    dizzy
    Participant

    I think it’s very worrying. Whilst advancement in technology must be seen as a good thing, redundancy for most human beings is a worrying concept,

    As an exercise try and list jobs that can only be done by humans, and will never or are unlikely to be replaced by AI There aren’t many.

    I think that caring professions which require human to human contact like nursing, the teaching of developing children, and perhaps wet nursing, might escape the worst of the cull, but there are not many jobs or professions which will be exempt. This will affect people right across the spectrum from manual workers to the traditional professions.

    A world in which virtually everyone will depend on the goodwill of the few super rich, not just for subsistence but for money with which to enjoy the fruits of the AI revolution is terrifying.

    More than anything I predict a catastrophic rise in mental health problems as people en mass struggle to come to terms with the fact that they are redundant. We are only glimpsing the thin end of the wedge at the moment.

    #3565

    Peter
    Participant

    @dizzy You shouldn’t be so gloomy!

    So far technology has created more jobs than it has destroyed but if technology finally does make us all redundant in the workplace then great, we can do the thing that we enjoy every day without work. People aren’t going to put up with having their lives ruled by corporations, in a democracy we can change things and I have no doubt that is what we will do if we need to. I see us all living in a kind of Bahrain where we all get a living wage and technology generates the money.

    However, I’m wondering which technologies you think will result in mass unemployment? We’ll lose a lot of driving jobs in the next couple of decades but other than that?

    • This reply was modified 3 weeks, 1 day ago by  utterz.
    #3566

    oldman
    Participant

    @peter I generally agree with your view over the doom and gloom views – ever since the Jaquard loom and the 2nd coming of the steam engine (the ancient Greek’s had a basic engine but never realised its potential, perhaps because it was hard to see it when you had a slave economy), technology has *eventually* created more jobs than it displaces, and has raised standards of living and life expectancy.

    But, jobs are lost and society and government hasn’t historically maintained those higher standards of living for those individuals on the end off a round of mass unemployment. There’s only so much retraining you can do. Generations have been blighted around me as a result of the lack of support for the end of most coal mining.

    I’m wondering which technologies you think will result in mass unemployment

    That’s the big question. The combination of EVs, autonomy and shared-ownership and on-demand business models will eliminate most petrol stations, perhaps 75% of work in maintenance garages, most ‘professional’ drivers, perhaps 80% of jobs in the automotive insurance industry, perhaps 75% of the jobs manufacturing and supplying after-market parts, perhaps 40% of vehicle manufacturing jobs, the list goes on.

    Farming will eventually be another mass casualty – one day most things will be grown in automated inside environments, I am sure – including cultured meat. This I think is further away than the vehicle stuff above. Mind you there’s already one firm shipping high-density hydroponic farms in 40 foot containers.

    #3567

    dizzy
    Participant

    However, I’m wondering which technologies you think will result in mass unemployment? We’ll lose a lot of driving jobs in the next couple of decades but other than that?

    @peter before I start listing them, can I ask what your job is (was)?

    #3568

    Peter
    Participant

    Before I start listing them, can I ask what your job is (was)?

    @dizzy Design engineer in the electronics industry, for the last two decades anyway.

    #3569

    dizzy
    Participant

    Design engineer in the electronics industry, for the last two decades anyway.

    Do you honestly believe that in the not too far distant future, maybe not in my lifetime, but probably in yours, AI will not be able to do what you do, and quicker and more efficiently?

    My list for starters:

    All driving jobs a you have identified, including flying aircraft and piloting ships
    All manufacturing jobs
    Professional jobs including Accountants, Lawyers, Doctors, Surveyors, Engineers, etc etc will decline

    Lets consider the Medical profession for example. Currently so much new research and medical advances, together alerts is published daily that no human being has the ability to keep up with it. Imagine all this constant data being obtained, stored, analysed, and acted upon by super computers which within micro-seconds can make a diagnosis and recommend treatment almost instantly? If a patient presents with a new unknown disease in the depths of the Amazon, it will be almost instantly available for comparison throughout the world. If a certain treatment regime fails and another succeeds, that information will become available throughout the world.

    Robotic surgery is already starting to take place. AI will be able to monitor patients before, during and after surgery adjusting calculating and adjusting anaesthetics and drug administration much quicker and more efficiently than humans can

    Consider Lawyers. Super computers will be able to assess all the facts, all Law and Case Law, and make a reasoned judgement or in micro seconds. Legal Opinions will be a result of considering many more factors than a human can, and much quicker.

    Engineers – masses and masses of data can be obtained and acted upon in design and structural calculations far quicker ans as accurately if not more accurately as humans can.

    Computers will be programmed to be capable of original “thought” as we call it, based on all the information and more that can be created by a human brain.

    Computers will design more powerful computers than themselves.

    Even in the world of Art I can see computers of the future creating new music, new paintings, new sculptures etc etc.

    Computers will be “trained” to out think the human brain.

    This is an interesting paper

    The AI Revolution: The Road to Superintelligence

    The AI Revolution: Our Immortality or Extinction

    #3572

    troll
    Participant

    Engineers – masses and masses of data can be obtained and acted upon in design and structural calculations far quicker ans as accurately if not more accurately as humans can.

    Just on this, what you say is already true. What computers can’t currently do, and I imagine will not be able to for the foreseeable future is make high level judgements. For example, is a suspension bridge or cable stayed be better here, taking account of a host of intangible and ill-defined considerations. Should the bridge go here or there or 35m that way? Should the design life be 50 years or a 100 years?

    #3573

    Peter
    Participant

    What computers can’t currently do, and I imagine will not be able to for the foreseeable future is make high level judgements.

    @troll Have a look at the links to Parts 1 and 2 of the paper on the Growth of the AI revolution which I edited into my previous post whist you were posting yours.

    • This reply was modified 3 weeks, 1 day ago by  Peter.
    #3575

    ben
    Participant

    I think there are a a few of separate issues with AI worth considering:

    First is the inevitability of automating our jobs, and us having to find new ways to do an economy. Good little clip here:

    Then there’s the questions surrounding making machines that are somehow human-like in the way they can communicate which either:

    a) seem so much like humans that we form relationships with them, even though they aren’t conscious; or

    b) we actually make something that is conscious and have to give it rights

    These are interesting questions, but we’re absolutely nowhere near making something like this. Look at the most advanced chatbots, e.g. the interview by the bot presenter on the Today Programme this morning…no one’s passing the Turing Test just yet! As for making something conscious, that’s ludicrous at the moment since we don’t know anything about the criteria of a system to generate consciousness.

    And then we’ve got the danger of the death of humanity to robots, which some people are losing sleep over.

    I think he’s talking nonsense, although it does sound fairly convincing the way he puts it. He doesn’t give any definition of general intelligence, and I think that consciousness may well be required to display general intelligence. To understand what’s going on in the world may require a different kind of information processing – that done by the brains of creatures but not by computers. So his main assumption that “intelligence is just a matter of information processing” may be glossing over an enormous, fundamental problem in creating these dangerous super-intelligent machines.

    #3577

    oldman
    Participant

    All manufacturing jobs

    @dizzy This doesn’t need AI, it needs robotics and machine vision. The reason meat bags are still used in many manufacturing jobs is that they’re cheaper than the robotic alternative. The price of robotics is falling, but in a pretty standard way for a maturing technology, the field isn’t on the cusp of an explosive development in the way many – myself included – believe AI to be.

    Consider Lawyers. Super computers will be able to assess all the facts, all Law and Case Law, and make a reasoned judgement or in micro seconds. Legal Opinions will be a result of considering many more factors than a human can, and much quicker.

    This I think will be truly transformative – a democratisation of law with everyone able to ask a computer system to adjudicate and explain. Mind you I doubt our written and case laws form a single, logically consistent corpus – untangling that knot will be fun.

    Computers will design more powerful computers than themselves.

    That’s been happening for 30 years. I’d hazard a guess that computers do 1000x more work designing a microchip than humans. I know just how much the EDA tools I used to use did “under the hood”. I don’t think a human has actually designed a microchip fully without computers for several decades.

    Computers will be programmed to be capable of original “thought” as we call it, based on all the information and more that can be created by a human brain.

    I am a firm believer that original thought is beyond computers, no matter how much bigger and faster we make them. I think this because I do not believe conscuiousness is computible (in a very formal sense). Mind you, given a source of non computational random noise, a computer can translate that into idea after idea for addressing some problem, and test them out against some criteria, identifying candidate solutions. Inspiration, if you like. But it’ll come from the randomness, not from the computer.

    #3578

    ben
    Participant

    Lets consider the Medical profession for example. Currently so much new research and medical advances, together alerts is published daily that no human being has the ability to keep up with it. Imagine all this constant data being obtained, stored, analysed, and acted upon by super computers which within micro-seconds can make a diagnosis and recommend treatment almost instantly? If a patient presents with a new unknown disease in the depths of the Amazon, it will be almost instantly available for comparison throughout the world. If a certain treatment regime fails and another succeeds, that information will become available throughout the world.

    @dizzy Computers that can synthesize factual information in this way will be extremely useful in medicine. However, until they can genuinely *understand* what a patient says, they will be a tool used by a human, rather than a substitute. In order to understand what a patient says, you have to grasp all kinds of subtle meaning under the surface of the words. You have to get a feel for who you’re talking to, the way they use language, when they’re talking literally versus when they’re using metaphor, and when they’re frankly just talking crap for some other non-factual, human reason that doesn’t feature anywhere in a database of symptoms and phrases. What we do with language is we express from one conscious mind to another the content and character of that mind. What a computer does is matches patterns of sound waves with words in a database which map to another set of symbols representing things that have meaning to humans. This process just doesn’t cut it when it comes to having a conversation – there is no reason to believe that without making the AI conscious it will ever be able to talk to a patient and interpret their words, their facial expressions, their body langauge etc well enough to generate the correct outcome.

    This does of course radically change the role of the doctor to one of communicating between the patient and the information processing machine that does the factual diagnosis and management once the correct input to this process has been determined. I think it may be a lot easier to train as a doctor in future!

    Even in the world of Art I can see computers of the future creating new music, new paintings, new sculptures etc etc.

    There are certain genres in which without training it will be difficult for an audience to distinguish the art made by a computer compared to the art made by a human. Very abstract, or very limited formulaic art can easily be made by computers. But a computer writing a song that connects emotionally with an audience? Absolutely not until one is made conscious. You can’t generate genuinely expressive content without understanding what is being expressed, and that understanding requires consciousness.

    Computers will be “trained” to out think the human brain.

    Too grand a statement. Without consciousness, all a computer can do is process information. Yes, computers can and will to ever greater degrees out *compute* the human brain (as the chess and go machines have done). But we have made no progress whatsoever on getting a computer to *think*.

    #3579

    robert
    Participant

    Ultimately we will all be replaced be technology as a combination of AI and present programmed machines.
    Driving we all know is doomed.
    Supermarket checkouts are 90% self service now, it’s only a matter of time before all shops are the same.
    A lot of manufacturing is done using CNC machines.

    The only thing slowing down the rise of the robots is the cost, people are far cheaper but the balance is changing.
    The tech is getting better and cheaper.
    Humans are getting more expensive and have more rights.
    The minimum wage keeps going up, employers now have mandatory pension contributions, working hours are lowering (hardly a week goes by without some think tank calling for 3 day weeks so that we have to employ double the amount of people to cover the same workload but people will still need paying for a full week to survive), maternity and paternity pay, holiday pay, sick pay, re stating holidays if someone is sick whilst on holiday, more and more onerous and expensive H&S legislation.
    Employing people is getting to be massively expensive.

    I’m just glad that my field is safe (for now) there are too many variables in on site maintenance for it to be robotised until 2 things happen.
    1 a robot that can do every movement a human can do and 2 designers re thinking the design of machinery to facilitate robot repair guys.

    #3580

    joe
    Participant

    @oldman It depends how you define original thought. What about Deep Minds alpha go? It’s now better than any human. There latest iteration learnt from playing and it’s self in a fraction of the time the old algorithm learned from past game training data. Some of the experts it played say it had developed completely new strategies they had never seen before.

    Regarding consciousness, I used to think like this too. If you look at many breakthroughs in scientific fields, the actual algorithm was developed from a hunch or leap in understanding that was later proved in some way, often with maths. Can computers ever do this? I don’t think we even understand how we do it. Is consciousness required for this? What is consciousness, does it emerge or is it a separate thing? Without consciousness to what extent can AI replace jobs (drivers etc). I think it’s an incredibly complex subject but in any case I think AI will replace a lot of jobs in the not far distance future.

    #3583

    Matty
    Participant

    @joe Original thought is beyond what we currently understand as computers. But then again one of my mother’s classmates back in the 60s declared that the limit of human invention had been achieved with the arrival of the pocket calculator. We may well oneday look back upon our current times as a primitive dipping of our collective toe into the ocean of possibilities. After all the silicon revolution itself is barely out of short trousers. Firm beliefs are all well and good but history is littered with disproved firm beliefs (although it seems that the 2017 view may be that the earth is flat after all??!).

    Not yet discussed on this thread is the so-called concept of singularity whereby AI will reach a stage when of its own volition it triggers runaway, unforseeable, change. The question to me is (barring our total extinction) why would this NOT happen?

    #3584

    nobroo
    Participant

    Computers that can synthesize factual information in this way will be extremely useful in medicine. However, until they can genuinely *understand* what a patient says, they will be a tool used by a human, rather than a substitute. In order to understand what a patient says, you have to grasp all kinds of subtle meaning under the surface of the words. You have to get a feel for who you’re talking to, the way they use language, when they’re talking literally versus when they’re using metaphor, and when they’re frankly just talking crap for some other non-factual, human reason that doesn’t feature anywhere in a database of symptoms and phrases. What we do with language is we express from one conscious mind to another the content and character of that mind. What a computer does is matches patterns of sound waves with words in a database which map to another set of symbols representing things that have meaning to humans. This process just doesn’t cut it when it comes to having a conversation – there is no reason to believe that without making the AI conscious it will ever be able to talk to a patient and interpret their words, their facial expressions, their body langauge etc well enough to generate the correct outcome.

    This does of course radically change the role of the doctor to one of communicating between the patient and the information processing machine that does the factual diagnosis and management once the correct input to this process has been determined. I think it may be a lot easier to train as a doctor in future!

    @ben I agree with much of the above but I’m not sure it will make qualifying as a doctor any easier, partly due to high levels of industry protectionism. Ditto the law and accounting sectors, despite views expressed upthread to the contrary. The big professional services firms are rubbing their hands with glee at the onset of new technologies, not shrinking in terror. Did the release of Sage Accounts reduce the world’s reliance on accountants? Or the spreadsheet, for that matter?

    Technology has a long history of creating employment, not destroying it. The issue is likely more one of anticipating the migration of work, geographically and by skillset, than of witnessing its decline. The picture will vary greatly according to locale and sector.

    None of the above is to suggest that, for those with less professional capital, this won’t be an extremely challenging revolution in work.

    #3585

    oldman
    Participant

    Firm beliefs are all well and good but history is littered with disproved firm beliefs (although it seems that the 2017 view may be that the earth is flat after all??!).

    @matty Oh I’m not denying that we’ll make machines with conscious-like and self awareness-like properties. Far from it – I’m almost certain it’ll happen.

    But it will be a ‘machine’, not a ‘computer’. A modern computer (including recent custom ‘AI’ microchips) are direct descendants of the calculators you mention – microchips that add, subtract, multiply, divide and compare numbers according to pre-written instructions, and that choose different collections of instructions based on the results of that maths etc. I don’t believe that conscious free-will can be created by such a system or can emerge from it. As Rom asks:

    Do we even know what consciousness is anyway? No – we do not. Until we do, recreating it is a bit of an inbounded problem.

    There is a mathematical theorem that hints that aspects of humanity can’t emerge from a computer, and that aligns with the subject of original though (see my next post).

    Perhaps I will be proved wrong – if so I won’t mind because the proof will be a phenomenal achievement that both changes the world and how humanity understands itself.

    Not yet discussed on this thread is the so-called concept of singularity whereby AI will reach a stage when of its own volition it triggers runaway, unforseeable, change.

    We’ve been having technological singularities ever since farming was invented, but this one could be the fastest one yet. Normally they’re multi-generational giving society time to adapt…. perhaps not this time.

    #3586

    oldman
    Participant

    It depends how you define original thought. What about Deep Minds alpha go? It’s now better than any human. There latest iteration learnt from playing and it’s self in a fraction of the time the old algorithm learned from past game training data. Some of the experts it played say it had developed completely new strategies they had never seen before.

    @joe Other AIs discover strategies as well – I like the ‘rattling’ in Breakout here – https://www.youtube.com/watch?v=cjpEIotvwFY

    Regarding consciousness, I used to think like this too. If you look at many breakthroughs in scientific fields, the actual algorithm was developed from a hunch or leap in understanding that was later proved in some way, often with maths. Can computers ever do this? I don’t think we even understand how we do it. Is consciousness required for this? What is consciousness, does it emerge or is it a separate thing?

    Exactly. The Go and Breakout strategies already exist as one of a finite number of possibilities in a system governed by strict rules that are easily discoverable through a well bounded set of experiments (moving the paddle left or right in Breakout…). A totally dumb, non AI exhaustive search would discover the same strategies, the AI just does it faster.

    Proper original thought is unbounded. Taking Einstein and General Relativity – almost a century of experimentation has since been going on to validate his work. It’s not clear to me that a bounded computer can discover new theorems in this way – to date they find things by chaining together existing proofs etc, but that doesn’t make the leap of imagination that delivers a radically new theory.

    Without consciousness to what extent can AI replace jobs (drivers etc). I think it’s an incredibly complex subject but in any case I think AI will replace a lot of jobs in the not far distance future.

    I don’t think consciousness is needed for many jobs, and I think it suffers in a lot of them as a result. The solution to a driverless car needed a conscious decision once every 10,000 miles (or whatever) appears to be to transfer control to a human operator in a remote call centre. That basically works now, and is ‘good enough’.

    There’s a done-to-death SciFi trope about what happens when machines achieve consciousness, but that doesn’t make it less worrying…

    #3587

    ratface
    Participant

    However, I’m wondering which technologies you think will result in mass unemployment? We’ll lose a lot of driving jobs in the next couple of decades but other than that?

    @peter White collar jobs. Specifically semi repetative Jobs involving number, data, spreadsheets and so on. Someone i know has started a machine learning company with a couple of other ex oil industy scientists and some developers and im very impressed. It won’t replace creativity interpretations, but it will save a huge amount of time and thus people.

    Going from that example i look at what banks, city firms do and if i was a CEO of one of those I’d set a target of 80% staff reduction in the next 10 years just because getting people to do routine data tasks is a horrible situation.

    Where it leaves society and specifically thr UK economy i dont know, but this i think, for good or bad is inevitable.

    If you want a wry smile it will soon be the case i think that all those people who did accountancy, business studies will soon be thinking id done something marketable like art, drama or media studies. There will still be many Jobs, tasks, but that middle class paper shuffling later of the service economy will be rather reduced . UBI?

    • This reply was modified 2 weeks, 4 days ago by  utterz.
    #3588

    don
    Participant

    @ratface The paper shuffling economy in the uk has for the most part already vanished to ……..Mumbai or somewhere similar.

    It’s too expensive to do in the uk .

Viewing 29 posts - 1 through 29 (of 29 total)

You must be logged in to reply to this topic.