Jump to content

The next great technology & change?


Recommended Posts

8 hours ago, cantab said:

That incident recently with the Tesla being "summoned" and crashing into a multimillion dollar airplane made me think about AI. While it's true a human might make that mistake, I think most people would be more cautious driving around an airfield, because we know planes are expensive. The AIs we currently have are being "trained" on narrow tasks, they don't have the lived experience that a person does, and I wonder if that might limit AI in some respects. It's not that Tesla's AI didn't recognise an airplane, it's that it doesn't know what an airplane is. All it knows is that there's stuff it's allowed to drive towards and stuff it's not allowed to drive towards. Just what level of safety can such a narrow mind have? There have been other examples of the Tesla autopilot veering towards cyclists and stuff.

To get an AI that can replace a human, might we not have to essentially raise it as a human? Is anyone attempting that? Would such an experiment even be considered ethical?

A problem would immediately arise of collecting a sufficiently large dataset, and then thoroughly cleaning it. Anything short of going through it with a toothbrush can result in the AI developing a weird bias, either unintentionally or, worse, intentionally (e.g. poisoning the data to tech the AI to ram any vehicle with Pennsylvania plates), it would take a lot of testing to even identify that bias, and then you'd have to start the whole process all over.

The problem seems that humans have a lot of built-in "brakes" that prevent us from doing something potentially harmful. We develop this over a huge period of time, and we're usually able to override impulses from an ill-trained neural net of our own. Whereas our AIs don't have that, and so, in the same place, if an AI decided it wanted a donut it plausibly could rob someone in the street for a donut if something clicked in its brain.

Our AIs are sociopaths.

Link to comment
Share on other sites

2 hours ago, DDE said:

Whereas our AIs don't have that, and so, in the same place, if an AI decided it wanted a donut it plausibly could rob someone in the street for a donut if something clicked in its brain.

Reminds me of those scaly things killing off all of humanity (thus all potential bearers of mutants) to “police the mutants” as instructed by the humans in X-Men.

Link to comment
Share on other sites

On 5/11/2022 at 3:55 AM, JoeSchmuckatelli said:

The examples given (largely focused on the US/West) are:

  • Factories
  • the Railroad 
  • Electricity 
  • the Automobile, and finally, 
  • the Microchip

Each new technology had a massive impact - a transformative impact - on not only the economy, but also society and how humans use our time and interact with the world. 

I know this is not a complete list, but I personally think the washing machine is missing. 

Imangine a world without the washing machine. How many extra hours of a day is needed to do the laundry? Albert Einstein will spend his time washing his shirt rather than thinking about the loophole in Newtonian Physics. Many men and women will stay in the house washing clothes instead of working to boost the economy and develop cutting-edge techonlogy. Without the washing machine there would possibly not exist Microchips or the ISS till today because their developers and engineers need time to wash their socks and trousers and their effective time for calculation is low. Also there would not be a game called Kerbal Space Program and this forum wouldn't exist. 

So the next game-changer may be something that lets a country pioneer. But it may also be something that benefits the entire population of 7 billion (including the current population sunken in poverty). That seems like a big word but maybe it eleiminates the need to do something completely mechanic and involves little creative thinking. What can you think of? A machine that can clean your skin/teeth/etc. automatically so you don't have to spend half an hour taking a shower and getting dry?

Link to comment
Share on other sites

10 hours ago, cantab said:

That incident recently with the Tesla being "summoned" and crashing into a multimillion dollar airplane made me think about AI. While it's true a human might make that mistake, I think most people would be more cautious driving around an airfield, because we know planes are expensive. The AIs we currently have are being "trained" on narrow tasks, they don't have the lived experience that a person does, and I wonder if that might limit AI in some respects. It's not that Tesla's AI didn't recognise an airplane, it's that it doesn't know what an airplane is. All it knows is that there's stuff it's allowed to drive towards and stuff it's not allowed to drive towards. Just what level of safety can such a narrow mind have? There have been other examples of the Tesla autopilot veering towards cyclists and stuff.

To get an AI that can replace a human, might we not have to essentially raise it as a human? Is anyone attempting that? Would such an experiment even be considered ethical?

And then there's adversarial inputs against image-recognition AI. Get an object, put a specially-designed sticker on it, and the AI is fooled into thinking it's a completely different object. It's a mistake no human would make. Of course we make our own mistakes, they're called optical illusions. But we live in a world designed for what human vision can and can't do. Machine vision that's unlike our own - which means ALL currently existing machine vision - will always struggle in that environment. Redesigning our world for the machines, well, it might happen some day, but I don't think that would be a good thing.

Ten years ago I though self-driving cars were just round the corner. I learned to drive feeling like I was learning an already-obsolete skill. But now, self-driving cars outside of limited geographic areas seem further away than they were in 2010.

Cars encounter planes very rarely so its not something its easy to train for, you also don't want the car to freak out about them. 
But in general cars should not drive into anything. It does not need to know that an piano is, just that it should not impact it. Same with shopping carts who it probably know about. 
And yes you can push shopping carts away with an car its not an AI need to do. 

Link to comment
Share on other sites

24 minutes ago, AllenLi said:

Albert Einstein will spend his time washing his shirt rather than thinking about the loophole in Newtonian Physics.

So, without the washing machine he could be doing two useful things at once: washing his shirt and thinking about the loophole in Newtonian Physics.

And would save a lot of electricty.

Makes sense.

26 minutes ago, AllenLi said:

Many men and women will stay in the house washing clothes instead of working to boost the economy and develop cutting-edge techonlogy.

So, twice as less people would be getting me with useless phone calls (the "sociological surveys", "tariff plans", and so on), and I would have at least a minute in shops to watch the goods on the shelves before a ninja in uniform cries in my ear "Good morning! What do you want?!"

Makes sense.

***

So, down with washing machines, go manual washing!

Link to comment
Share on other sites

I listened to a podcast a few years back which looked at time use surveys and concluded that washing machines made much less difference than you might think, basically people wore clothes for longer before washing them, used detachable collars, dark colours to hide dirt etc.

The thing that really made a difference was pre prepared food. The average food preparation time for a household in the 60s was around 4h a day, now it is 45m.

We were willing to be a bit grubby but not too starve.

The BBC did a great series called "50 things that changed the modern economy" that I had to re-listen to to write this

 

Link to comment
Share on other sites

One slightly different approach to "the next big thing" could be an overhaul to our IP laws.

Currently, the whole system makes a mockery of the whole concept of "supply and demand" as there is an infinite supply of each "IP unit" that can be replicated as needed to nearly everyone.  But modern economics is built on artificially limiting that supply, and typically intentionally damaging (see any Intel CPU, or most microelectronics in general) said product to allow different prices for segregated markets (another example is that it is cheaper to make "loaded" cars with all options than to build "base" models without them, at least until you add leather interiors).  Compare the price of "officially licensed" and "legally branded" products to their exact knock-offs (often made with less volume efficiency) and you get the idea of the price of all that IP.

On the other hand, alternative methods are few and far between.  Copyleft is making progress, largely because software is the most extreme example of the above effect, but rarely directly provides the support for programmers and other IP creators that the corporations hoarding their IP can.  Purely socialist economies haven't done so well, but it isn't clear that a system that privatized commodities (or production in general) and socialized the IP might not work.  It would run into tremendous international legal issues (to the point of banning discussion as politics), so could only be done by a country capable of working  outside of the current international legal system (through either ostracism or political power/size).

I really don't think 3d printers could niavely perform such a feat, although perhaps something like kickstarter and similar could evolve to match "microfactories" that print molds and whatnot and quickly do small runs of specific products would be a thing.  Getting crafters to figure out "all the steps in between" would probably be interesting, and lead to even more gripes about failed projects than currently on kickstarter.   It is certainly a far more political/legal issue than a technological one, although certain technologies (like software is now) might speed up the process. NOTE: an idea very similar to kickstarter was floated on rec.arts.sf-lovers way back when the internet was new and established authors thought the whole idea was crazy that people would fund sequels to otherwise unprofitable works that were a favorite of a small amount of readers.  Or perhaps Patreon would be a better model if relatively few people shouldered the burden of getting otherwise unprofitable books written and published.

Link to comment
Share on other sites

2 hours ago, cubinator said:

I think the pandemic has catalyzed a bit of a societal shift. I think maybe the thing to come first might be a change in the way we conduct ourselves, before any particular object.

Has it? I'm not that sure. Barely three months after an omicron wave, I'm seeing undiminished crowds, next to no masks (<1% use on public transport), and the subway is removing free sanitizer stations. People are in a rush to return to normalcy. The only holdout I can still see are people (not) at work, and that's only because of a very clear lack of need for in-person attendance in my industry - and because the office admin is yet to return all the chairs that they've taken away to enforce social distancing.

Link to comment
Share on other sites

4 minutes ago, DDE said:

Has it? I'm not that sure. Barely three months after an omicron wave, I'm seeing undiminished crowds, next to no masks (<1% use on public transport), and the subway is removing free sanitizer stations. People are in a rush to return to normalcy. The only holdout I can still see are people (not) at work, and that's only because of a very clear lack of need for in-person attendance in my industry - and because the office admin is yet to return all the chairs that they've taken away to enforce social distancing.

I say "catalyzed" because I'm aware it's not a total upheaval, nor is it a complete reason for any future society being the way it is. It's an experience that moves things slightly in another direction for the future, even in the case you've observed of people working from home more. That and other pandemic-related experiences will be an influence on decisions people make from now on.

Link to comment
Share on other sites

Posted (edited)

as far as great plagues go, this one was kind of meh. think of all the changes that happened because of the spanish flu (i cant think of any). thats about what we will get out of this pandemic. 

the only thing different this time was we had biotech to roll out a vaccine in less than 2 years. just goes to show there is not a virus on the planet that cant be nuked from orbit provided you want to spend the 2 years and the millions of dollars it took to develop the vaccine. that really doesn't save us from future pandemics, unless we can make the lead time shorter. 

Edited by Nuke
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...