The international members

(pip linney-barber) #21

It’s as though the technological revolution, that has obviously and profoundly changed everyone’s lives, didn’t really happen. ‘Apprenticeships and traineeships’ for what? I’m sure the Trades Hall have decent intentions but…2018! Not 1948.

I could be wrong but as far as i’m aware no political party is talking about the reality of an automated world. Pirates ought to take the lead, i agree John.

(Andrew Downing) #22

I’ve been a software engineer for ~35 years, and I’m currently working in a research team that is just full of people who specialise in machine learning (the basis of AI), so I see what’s happening from the inside of this technological trend.
Here’s how it looks to me.

The biggest job under imminent threat, is “Driver”.
There are literally billions of dollars being thrown at self driving cars by probably half a dozen multi-national corporations, surrounded by thousands of companies wanting to get even a tiny slice of the pie, in what is estimated to be something like a $15 Trillion / year industry globally.

There are self driving prototypes driving around on the roads already, but under human supervision.
There’s an international SAE standard that defines standard levels of self driving capability:

  • Level 3 is where a vehicle can self-drive autonomously on any normal road, but in the case of a situation it doesn’t know how to handle, the human is the fallback. You’re expected to be prepared. There’s a lot of people trying to figure out better ways to ensure you will be prepared to do that.
    Level 3 is already working.
  • Level 4 is where there is no requirement for human fallback, but if you get off the beaten track, you could optionally still drive it yourself.
    Level 4 is will probably be realised in the next 4-5 years.
  • Level 5 is where there is no option for human control any more.
    Level 5 is not really harder than level 4.
    It’s just that vehicle designs change. People become cargo.

This has a lot of weird implications, some of which are less than obvious.

  • The really obvious one, is that given that “driver” is the most common “unskilled” (or at least common skilled) job, there are a lot less fall-back positions for people that can’t get other more specialised jobs with higher pay.
  • Less obvious: People are going to stop owning personal cars. They will just pay for the use of a vehicle, that will come an pick them up and take them wherever they want to go. If what you really want is some lumber from Bunnings, then you don’t really need to go there, an appropriate vehicle can just go get it and bring it to you then go away again.
  • Less obvious: Most vehicle servicing will become centralised. Say bye bye to your friendly local service mechanic. There will be a lot less but larger service centres, and the cars will mostly take themselves there.
  • Less obvious: Car insurance will radically change. At level 4, vehicle accident liability will be with the vehicle manufacturers, who will build that into the price and they will have a strong incentive then to keep making them safer, but they also get to re-insure that and take more margins. That liability factor will also act as a barrier to entry for any would-be small makers of self driving vehicles.
  • Maybe obvious: The licensing, insurance and ownership costs and conditions for anybody wanting to manually drive a car will go through the roof.
  • Probable: Public transport gets a whole lot more useful. Imagine a little vehicle that comes to pick you up for work, then joins a whole train of other little vehicles to efficiently roll along together into the city, then splits out to drop you off at the office. There are already designs floating around that plan to operate like this, but they probably need to wait until most human driven traffic is out.

In this same time frame, the vast majority of vehicle manufacturers world wide have already switched to focus all of their R&D on electric cars.

  • Maintenance is massively lower work and lower cost. The motor mechanics end up with a lot less to do.
  • Most of the consumable parts are gone. Tyres are similar, brakes wear out less because regenerative braking, no cooling fluids, no petrol, no oil etc.
  • The power infrastructure to support this is going to have to grow massively.
  • We’re going to need a lot more electricians. Ramp up those apprenticeships

The other big change I see bubbling along in the background is automated house construction.
There’s a lot of different approaches being tried:

  • 3d House Printers - like a gantry system on your block with a concrete nozzle that prints out walls or lays bricks under computer control for whatever house design you enter.
  • Factory pre-fabrication of complete walls for designed-to-spec houses, that are just connected together on-site with a crane and a few tools. Instant house in a week, with all the fittings, electrical and plumbing built in.
    Either of these do away with much of the existing building trade roles though houses should get cheaper. It’s hard to predict the timing of this change.

Another big change is automated fabrication, of just about anything.
3d printing (or “additive manufacturing”) and CNC Milling (or “subtractive manufacturing”) under computer control is shaping up to be a really big deal, but perhaps weirdly, assembling things is still a pain in the arse.

  • You can 3d print in just about any material you like, including titanium and steel by using laser sintering (the patents on most of this recently expired so there’s explosive growth in this area just now) .

(Andrew Downing) #23

Well that was a bit of a tangent.

I could see us selling the general Australian public on taking some more forward thinking approaches to the constantly encroaching technological changes that they see all around them.

I can’t really see how we’re going to be convincing about the killer robot scenario.
I don’t even think it’s particularly realistic. Fully autonomous killer robots will be vastly more of a liability than an asset to any military force, for quite a while yet.

They can do things like getting robots to run over rough terrain, or fly through difficult spaces, but expecting them to understand complex unstructured environments, where they’re expected to make sensible strategic decisions is just not effective. Using deep neural network techniques that are the typical approach to this sort of thing still actually means that when it does the wrong thing and shoots the wrong person, they won’t even be able to figure out why, A great example of this happened a couple of years back, where Google put out a fairly general visual recognition system for people to play with, but ended up very embarrassed when it mis-identified some African American folks as gorillas. It’s all a bit machine-see, machine-do at the moment.

The military will stick with just killing people the old fashioned way for quite some time to come I expect.

(Andrew Downing) #24

@John_Wilson , I’m still really hoping to get your union-experienced take on this strategy.
I may not have explained myself well. Not sure.

(pip linney-barber) #25

And the more white collar jobs? Accountants for example and legal professionals might also face job losses with more sophisticated software. Bio-technology could reduce our need for doctors and surgeons too. Bank managers thinking about loans? Why not just run the algorithm over the applicants entire life and medical history, which is all online.

I sometimes have to stop myself and wonder if this is some kind of dystopian conspiracy theory but it’s kind of already happening. Thing is, it doesnt have to be dystopian. If we can run a vast lucrative economy with not many people needing to contribute its, at least economically, just a matter of redistribution.

The broader challenge is meaning, but that can evolve. We need to break out of the job cult. You do not have to have a job to live a meaningful life, unfortunately this notion appears unfathomable to many people. This needs to change.

(pip linney-barber) #26

Perhaps a role for unions might entail helping former members (workers) to negotiate a jobless life. That task will need significant investment. How to build communities not based on jobs. How to make friends, not based on the workplace. How to feel useful and meaningful without a job. This is a massive cultural, even civilisational, shift that needs careful thought and care. But it can be done, i hope, i suspect it will have to be done.

(Andrew Downing) #27

Not sure what it means to “negotiate a jobless life”.
Who would you be negotiating with?

The cultural shift is going to be more problematic still.

I think this is a MUCH deeper problem.
I think that as humans, we need to take on some kind of responsibility in our lives, or else we just become nihilistic and depressed, then we wither away and die, one way or another.
Purpose matters. A lot.

(John Wilson) #28

Andrew. I suggest you dig a bit deeper and no I am not off on a tangent. Population Surveillance & Population Control is a big part of this gig. Research it.

Don’t be so melodramatic re: trying to sell a killer robot scenario ti the Australian Public. I never suggested this and don’t know how it became part of the conversation.

Insofar as inequality, and job security etc, well this all becomes redundant, kaput. It is a paradigm not based on reality.

(Andrew Downing) #29

Well, you wrote this:

Maybe I badly misunderstood you but that read a lot like a killer robot situation to me, and then followed by:

Robots “unleased against humanity”, by “psychopaths in the military industrial complex” and “articulate this threat to Australia’s Working & Middle Classes”.

Please clarify.

(pip linney-barber) #30

I’m using ‘negotiate’ in the sense of ‘finding a way over or through (an obstacle or difficult route).’
Just to clarify.

(Andrew Downing) #31

Ok @PLB. Kind of an internal negotiation then.
Makes sense.

(Andrew Downing) #32

I’m less worried about typical office jobs. They’ve already had their automation revolution to a large degree.

Lawyers is a different story. That profession has resisted change, but I can see a tipping point coming on, especially at the para-legal level. E.g. Machine learning systems to instantly find case law precedent etc.

Medical is also kind of different, in that it really involves human interaction. Every new medical advance seems to create a new function to be performed with another human in the loop. That’s why the total cost of medical thing keeps going up so much. It’s very open ended as a problem space.

(Andrew Downing) #33

The robots are getting quite impressive though:

Worth keeping an eye on what they’re doing at Boston Dynamics, and there really is a debate about giving these things guns.

(Alex Jago) #34

The key problem is not that jobs will be automated out of existence. In a system dominated by the interests of capital, the real risk is that technological change will further concentrate wealth and power in the hands of the dominant elite often referred to as the 1 per cent. As [Tim] Dunlop says, radical responses are needed.

John Quiggin, Inside Story, today.

(Andrew Downing) #35

Hey Alex, I was hoping there might be something like a strategy in that article. They suggested something along the lines of people owning their data and having rights to income derived from that. Problem is, that they didn’t bother to do the math.

Say for instance that we ballpark that Google makes $100 Billion per year from ad revenue. They have at least a billion customers, meaning $100 each. That’s not going to solve anybody’s income problem, and it costs a lot to run their massive infrastructure, so you ain’t getting it all.

We need better answers.

(John Wilson) #36

All, John Quiggin has pricked and opened up a key issue that is central to the concern held by many Activists such as myself. None of us are talking about how to stop AI nor Robotics. None of us are in any way deluded insofar as the introduction and proliferation of AI & Robotics. All of us are cognisant of the great public interests that could be served by this new technology.

But who owns it?

But who controls it?

But is the roll out going to be done for the well being of humanity or to enslave humanity?

When we start asking and answering some of these questions, then I will tell you if I support or reject the proposition. Thanks.

(Andrew Downing) #37

The default answer, is that the people who own and control it will be based on a tension between the people who make it, the people that buy it from them, and regulators. The well-being answer will emerge from that tension over time.

Slippery answer I know, but it’s correct. The regulatory remedies are the same old things Pirate Party has been campaigned about for years. Privacy, transparency, digital rights, etc.

I think the problem comes with trying to sell the significance of those things to Joe public. They don’t get it, or we haven’t stumbled upon the key messages that will draw their attention. It’s all too abstract for most people.

(John Wilson) #38

So say you.

Not I.

(Andrew Downing) #39

This isn’t going to work very well as a guessing game.

(miles_w) #40

Some research grants carry stipulations about public ownership, usually after a limited duration patent monopoly. The majority of private research is locked behind patents though with funding traced back to large international corporations. Some of the world leaders in automation and robotics include Boeing, Google and Amazon.

Traditional business mentality means those patents will be locked up behind protectionist agendas for decades without either regulatory intervention from a tech informed government or a tech company deliberately releasing their patents to the public, which Google and other companies have done in the past.

I don’t think we should wait for publicly traded companies to release future-critical RND products to the public domain. We’ve seen what decades of trickle down economics have done for the west.