ChatGPT - artificial intelligence.

[Spectric]

and told us that his lecturer said on the first day that everything they learnt in year one would be obsolete by the time they graduated, and their degree worthless within 5 years of graduating if they didn't continue to learn and study after graduation.

This has always been the case in electronics, technology has moved at such a pace that you had to embrace a process of continual learning in order to keep up, only the core theory of electricity and magnetism remains fixed. Computing will eventually reach a point of massive change once someone comes up with a viable alternative to binary, imagine a processor that instead of millions of switches with just two states you have something which can handle a dozen states.

 

[JimJay]

It appears that as a species we just never learn to see, let alone answer, the biggest question. Just because we can do something, does it necessarily mean that we should?

 

[XTiffy]

When I started in electronics and early electronic computers in the mid 1960s, the most often used phrase was "Rubbish in = Rubbish out". AI depends upon what is "put in" and by whom. One of the most over used words these days is "Expert", who picks them and who vets them? I would guess that the average "reporter or journalist" uses the internet and will quote based on their own political bias and will apply the term "expert" to whom they wish without any reliable checking of relevant qualification. Too often today we see cases of individuals who have faked qualifications and have got away with it for many years, sometimes in medical positions. Just who is going to vet AI? Give me paper based exams any day they are a true test of what has been absorbed rather than "looked up" which often applies to course work.

 

[nickds1]

This has always been the case in electronics, technology has moved at such a pace that you had to embrace a process of continual learning in order to keep up, only the core theory of electricity and magnetism remains fixed. Computing will eventually reach a point of massive change once someone comes up with a viable alternative to binary, imagine a processor that instead of millions of switches with just two states you have something which can handle a dozen states.

My degree was electronic engineering, physics and maths in the late '70s/early '80s.

As you say, all the core theory is valid, as is all the analogue theory, which whilst not so fashionable in this predominantly digital age, is vital for signal transmission and conditioning.

To be blunt, analogue engineering is HARD to do well, so folk tend to avoid it. RF is the same, and arguably an extension of analogue.

I have seriously enjoyed keeping up with developments in electronic engineering over the years - really interesting. I still design hi-fi valve amps, but also micropower IoT motes for remote environmental monitoring where every uA has to be accounted for. Electronic Engineering is fascinating - such a huge field.

 

[Spectric]

To be blunt, analogue engineering is HARD to do well, so folk tend to avoid it. RF is the same, and arguably an extension of analogue.

Yes I would agree, I remember doing a module that was all about transmission theory, modulation, side bands and electromagnetic spectrum etc etc and the maths was horrendous. I think a lot of people don't realise that they really live in an analogue world and don't really grasp the true meaning of digital which is so often used out of context. I grew up with valves, lovely things that give a nice warm look to anything they are used in, amazed when things like the OC71 transister appeared that supposedly could do the same job !

I to enjoy keeping up with changes in electronic technology but do find there are far fewer active people on some of the old forums and some of the newer documentation leaves a lot to be desired, but solving embedded C programs keeps the old grey mater active. One of my bigger problems is when prototyping circuits, far less through hole components, everythings shrunk versus being older with not as good eyesight.

 

[Molynoox]

Those images are certainly interesting, puts me in mind of game graphics some years back. If you got to the playable limits the physics got a bit wobbly. Can't imagine it will take long for AI to catch up if it is fed enough information.
The thing I find most intriguing is whether/when AI will be able to make 'intuitive' predictions by applying existing information to novel situations

Yes the resulting images were fascinating, not because they are useful, but because they indicate a certain trajectory with the technology. It's something I feel I want to have an awareness of because it will impact us sooner rather than later, certainly within my lifetime I would have thought.

I think you are spot on with your definition of utility as that is pretty much the backbone of what is expected of a scientific theory - if it can make predictions then it's good, of it can't then it isn't. Nice and simple but not for everyone perhaps as it flies in the face of a lot of other belief systems.

Martin

 

[artie]

Artie - could you expand on that thought please? Genuinely interested as to what you are thinking. AI related.

It's very hard to do without spending more time than I want to for the sake of a forum post, but a few things that come to mind are driverless cars, supermarkets without fixed checkouts, online doctor diagnosis etc.
These few things alone have the potential for many unexpected consequences.

AI could be a blessing because it shouldn't get tired after a 18 hour shift.and function perfectly, but then again we've all seen the digits crash.

There's also the ever present open to abuse.

Remember when the Parish Priest was almost god and what he said was law in the village.

Once people get used to asking AI what to do, it could be our new God

 

[aet1]

Yes but you could say that of many students struggling to cobble together an essay!
Given the right question would it recycle nonsense from the loony right (e.g. Jordan Peterson, Ayn Rand etc)? And would we be able to spot it?
Could be like sat nav which churns on intelligently giving out directions and suggestions, until it loses the plot completely and you are going around in circles?

I lost the plot after 18 years of teaching.

 

[Rodpr]

I have long been impressed by the way the internet has allowed knowledge to be shared much more easily than in the past. It is wonderful that people take the time to show others solutions to shared problems - who needs a Haynes manual when googling a problem nearly always leads to a video showing how it can be tackled?).

As a child I loved going to the library and coming home with armfuls of books full of fascinating information (and pictures) - hugely expanding the range if what I was able to experience. Today it is so much easier to find the information we want or need but unfortunately that comes with the risks of manipulation and deception which we have seen, for example, in the ways FaceBook was used to manage opinion ahead of US elections and the Brexit referendum here.
Nothing new about this, though. Wealth has always brought the opportunity to manage and 'correct' the information that people are able to access (e.g. ownership of huge media organisations and corruption of politicians). Despite this, I believe, easier access to more (and more diverse) information has got to be a good thing. Medicine is improving as doctors begin to realise that it is not enough to assume that what works for one group will work for all and, for example, that depression may mean very different things for different individuals, requiring different kinds of treatment (and more sensitive diagnostic processes).
The new AI tools help to distil what we need out of what other people have put into the public sphere and I am sure huge amounts of money will be invested in trying to control how information is accessed.

 

[Molynoox]

But is it, it is certainly a human organ that we really know little about in the grand scheme of things and is definately not just a collection of logic cells processing boolean algebra.

Yes I think the brain is a machine... just a very complex and capable one.

My question has always been how can you get AI to mimic a human brain that we don't even fully understand and if you have AI that is just pure logic, that could be a very dangerous route to take if every decision is just black and white. Now you might think that taking out the human emotions will improve efficiency in decision making, it would but just imagine a world without human emotions where people caring for each other is no more. At the moment processor based systems just respond to data provided by various sensors / devices and can be extremely precise, but if you think of making a robotic human using AI then some big issues, how do you program love, grief or happyness into a computer !

An AI of the future with human level intelligence may not need emotions to achieve the same results. If we wanted to exist symbiotically alongside AI, once it gets sophisticated, then we would need to ensure it has a similar moral map to ours, and that may or may not involve some sort of emotional overlay, but I suspect it wouldn't.

If / when fully developed AI becomes a reality this would be the point of a role reversal, instead of the dumb computer being a servant doing repetative task AI would soon take the lead, capable of digesting vast amounts of data and making decisions that could lead anywhere. We are already starting to sow the seeds for change with the internet of things, IOT in that we are connecting more and more systems onto networks and providing remote control from multiple locations so albeit slowly we are moving in that drection and how long before AI would determine that humans are actually pointless, or only fit to serve AI ?

Yup, scary isn't it. Humans may be the bootloader for AI as the next logical step in evolution.
Once AI reaches human level intelligence it will be able to design better AI systems and won't need humans for that at all. Can you imagine the speed at which it will be able iterate different and slightly improved versions of itself when new models can be built and tested 'inside the machine' in a matter of seconds or even milliseconds?

Once it reaches a certain point the speed of advancement will be almost infinitely steep and huge leaps in performance could happen as fast as you can click your fingers

Martin

 

[Peri]

Mis posted

 

[XTiffy]

As I have said before (in other post topics) being a cynic, I don't go all misty eyed thinking of the future of AI. Many years ago I can remember that all legal judgements etc. would be available to all and searchable. What happened to that project? If there was a case for AI the legal world is it. I don't expect that those profiting from legal work will allow that. We have seen more and more control of what we can learn and read (Facebook, twitter et al

 

[Peri]

Considering the end product that "teachers" are apparently content to be churning out these days, one might think that the questions should have been along the lines of how to do their jobs better.

Most teachers I know - and I work in education - are absolutely not content with the end product that gets turned out at the end of the academic year.

They are however totally fed up with:
Departments that have had staff cut by 75% whilst student numbers stay the same.
No wage rises for over a decade.
Effective teaching hours cut, then cut again because the staff simply don't exist to teach the hours that are really needed.
Having to basically teach to the exam/assessment because there's not enough time to teach the underpinning knowledge.
League tables that mean you might as well get a memo from management saying 'To show we've improved and to beat the college down the road we need a pass rate of 97.4% this year - make sure it happens'
Having to give apathetic, undisciplined students who don't want to be in education passes for exams they didn't want to take, just to satisfy the above point.

Fixing a couple of those points would help teachers do their jobs better.

 

[Jacob]

Yes I think the brain is a machine... just a very complex and capable one........

Yep biological machine. It's "consciousness" which worries some philosophers - but not all!

 

[Spectric]

I think Jacob has got it, stop thinking of computing in the electronic / digital sense and re think in the chemical domain.

 

[Dibs-h]

A friend of my son started a computing degree a few years ago, and told us that his lecturer said on the first day that everything they learnt in year one would be obsolete by the time they graduated, and their degree worthless within 5 years of graduating if they didn't continue to learn and study after graduation.

During Uni open days a few years back I took my son along to 2 Uni's to look at Computer Science courses. To say the lecturers were full of dung is an understatement - more interested in the useless crap they could write research papers on (and get back slapped by their peers) that were of no use to man nor beast.

I even asked one student (who'd done 2 out of 3 years) what the difference was between POP3 & SMTP and got a blank look.

I came to the realisation that what Uni's teach on the Computer Science courses bears little resemblance to the commercial world. As for "obsolete in 5 years" - that particular lecturer is chatting the proverbial. Languages like .net and React.js are well over 5 years old and still very mainstream, and design patterns like MVC are as current as it gets, even tho they first came to light over 10yrs ago.

I told my lad to skip Computer Science and study Engineering.

 

[artie]

I told my lad to skip Computer Science and study Engineering.

My lad took computer studies, I advised him to get a trade, plumber/electrician.

Almost 25 yrs later, he's managing a hardware shop.

 

[Spectric]

Is that computer hardware or hardware like paint and step ladders.

 

[Terry - Somerset]

The purpose of education must be to enable young people to develop the skills, behaviours and knowledge to function successfully in adulthood.

Traditionally this was tested by examinations under controlled conditions and mainly focussed on their capacity to regurgitate facts (history, geography, chemistry etc etc), communicate mainly through the written word, do sums etc

Exam boards asked the questions, students attempted answers.

In the new world the capacity to answer factual questions is no longer testing. Giving opinion on most moral or ethical issues is a little more challenging but there is little that one may reasonably ask of a person under 18 which has not already been copiously explored.

The real future of education is teaching kids to ask the right questions.

 

[Jacob]

..........The real future of education is teaching kids to ask the right questions.

And how to find out the answers - which would be "cheating" in an exam. Hence continuous assessments.
PS and how to handle that the answer might not be the one you want to hear, as with political threads, such as this suggestion from an eminent economist Joseph Stiglitz: tax high earners at 70% to tackle widening inequality