I'm not at all conviced. They point out minute, small progressions in the direction of automation and then extrapolate that to 'the professions we're so used to having will wither and die because AI, that's why'.
- Thus we turn up at our doctor’s with more web-collated information on our persistent leg wound than a field paramedic. If search engines struggle to turn up the answers we seek, we find the solution ourselves in a jungle of user forums where we interpret and judge the practical opinions of others. This amounts to the devolution of a classic professional competence.
What, so because I can read WebMD I'm suddenly in the same ballpark as a doctor with at least a decade of education behind him? Some people can figure some medical issues out sometimes. Maybe your GP has slightly less people with stupid questions. Doesn't mean the whole profession is devolving.
- And more students sign up for Harvard’s online courses in a single year than have ever attended its Massachusetts campus.
How many actually finish that and use it for something in their work life? Shockingly little. This is also implying that teaching as we know it is doomed because some MOOCs do reasonably well. Besides, since when is teaching is only about giving presentations?
- The Vatican has even launched an official app to help sinners prepare for confession (though with the usual proviso that it is no substitute for the real thing)
sigh. Pack it up religious leaders, it's time to go home, there's an app for that now.
Perhaps the book is better than this, for example discussing discussing the inherent complexities in those professions and the difficulty of automating them. The hand-waving that happens here is awful, though.
In my profession, it's necessary to simulate the performance of the electric grid. Forty years ago, such simulations were done by rooms of PhDs who collectively spent hours to produce one second worth of time domain results on a system composed of a dozen or two nodes.
Today, I alone can produce similar results with my meager bachelors degree on a system of 100,000 nodes in the time it takes to get a cup of coffee.
The article's negativity seems to start from the assumption that professions cannot survive change, and then takes a linear interpolation out to show their ultimate demise. Things have changed before, and while job descriptions change (I couldn't use a slide rule to save my life), people keep adapting and moving forward.
I think C.G.P Grey's "Humans Need Not Apply" video was a cultural moment. For people who had been following general singularity theory for some time all this is not news, but now that the wave of technological innovation is disrupting basic functioning of civilization in regards to -- I think the key categories of disruption include labour, property, and state -- now it will affect the way people in general imagine themselves and their relation to society.
Totally agree. Have you listened to the other conversation in that post? it goes into a lot more detail.
I think that this is basically the best argument for "mincome" or something like it. There are, in the not too distant future, going to be people who aren't just unemployed, they are unemployable. We will be in a "post-scarcity economy"
- Have you listened to the other conversation in that post? it goes into a lot more detail.
I know what I'm going to do tonight! Thanks for the link :)
- I think that this is basically the best argument for "mincome" or something like it. There are, in the not too distant future, going to be people who aren't just unemployed, they are unemployable. We will be in a "post-scarcity economy"
I, of course -- (I am not a monster!) -- I support Universal Basic Income (UBI). But I think the whole notion and societal conversation of a UBI suggests to me that there is something far more serious and fundamental that is wrong with our socioeconomic system (not that that is news...) and that we need to go beyond money, whatever, post-money, eventually. I think eventually it will be unnecessary and there will be a more effective and humane ways to build trust and collaboration between humans and the networks of AI/AGI that emerge throughout the century. Even sharing economy and gift economy models of organization are surprisingly varied and abundant, I have been looking more into different entities that function without money, and there are many of them! One of my favourite examples is, and I have used it quite a lot now, is couch surfing. It totally functions, it could of course function better, but it totally functions just on altruism and reputation systems, which enable a type of social trust without any money exchanging hands. I think there are signs for optimism here.
Do you really see money as the bad guy? Money is quite popular, and has enjoyed a high approval rating for some time.
mk mentioned that paper showing that money is really just a form of credit. People ran with it, concluding that banks manufacture money out of thin air, but the real lesson is that all of us have that power. When I forget my wallet and borrow lunch money from a coworker, my IOU is "money" in just the same way.
As is true of any powerful tool, money can lead to harm. (And is true of every good intention, trying to prevent such harms can lead to harm.)
The sharing economy has advantages for shareable goods, and already functions alongside the money-based economy for consumable goods. What if I sleep on your couch today, with the understanding that you can sleep on my couch some other time? To the extent that our agreement is a real obligation, we have just created a kind of money. If we have friends in common who enjoy mutual trust, you could perhaps exchange your couch ticket for a meal or something else.
If I want something, I can approach a complete stranger who is willing to provide it, almost anywhere in the world, and give them some colorful pieces of paper or their digital equivalent in exchange, and everybody walks away happy. What more effective and humane way to build trust and collaboration could be conceived?
- And is true of every good intention, trying to prevent such harms can lead to harm.
In a recently published book that I'm currently reading, Other People's Money, the author (John Kay) makes the point that heavy regulation is actually a big part of the problem in the finance world. According to his analysis, most of what investment banks are engaged in these days is arbitrage, which take the form of fiscal, accounting, and regulatory. Essentially, he says that governments and investment banks are in an ever increasing arms race (JP Morgan for example employs tens of thousands of regulatory compliance officers) that ends up being very profitable for the banks at the expense of taxpayers and borrowers, as well as entrepreneurs who can't hope to keep up with tax codes that take 1000 people to read fully.
Kay, although a "liberal" thinker by current popular definitions, suggests that much financial regulation needs to be scaled way back and modernized, and that "reform" is needed far more than regulation. Although I haven't reached his chapters on reform yet, so I'm not sure what he means by it exactly, I believe he is alluding to cultural biases toward making money at all other costs. He quibbles heavily with main stream economists' appropriation of the word 'rational' as only meaning "what makes the most money in the term we're currently concerned about."
I have probably disagreed with you more than anyone else on the site, but now and then you say something with which I could not agree more.
- an ever increasing arms race ... that ends up being very profitable for the banks
When we see social problems, it's common to expect government to make it better. That's their job, right? That may even be their intention, but intentions and results are two different things.
Responsible regulators know that it is possible to mess things up by interfering. To tread carefully, they might seek advice before acting. Who do they seek advice from? The experts, of course: the largest, most-connected industry insiders. (Sometimes the regulator saves a step by being an industry insider.) Even if they do not seek advice, special interests have incentive to lobby for favorable legislation, while the negative effects are too broadly spread to attract organized dissent.
The result, too often, is regulation that benefits the best-connected insider, at the expense of their competition, customers, taxpayers — everyone else. So when you suggest that we regulate marketing drugs, regulate wages, regulate real estate investment, regulate corporate finance, regulate Amazon, regulate farming, regulate legal financing, regulate horse shit, you may well be advocating for the entrenchment and perpetuation of these special interests and their unfair advantages.
Regulation or not, the thing we mainly need to keep in mind is outcome. I just had a conversation with mk earlier today about how frustrating it can be that NIH typically seems more concerned with mechanism than outcomes in what they choose to fund. Same applies to regulation. There can be good ones and bad ones, but what should define either isn't how "fair" the playing field is, but what the end product looks like. Imagining the world we would like to live in then thinking of ways we can make the world conform to that vision is a better strategy than ensuring rules are strictly adhered to.
'No one is indispensable' now arrives with a vengeance.
Not to worry. They can do anything so much better and hugely cheaper than mistake prone humans, we'll eventually be living high on that piggy for nothing.
Medical diagnosis was one of the first successful applications of expert systems, and has kept getting better. AI still doesn't have a big role in medicine, and likely won't for some time, because we want a human to blame when something goes wrong.
Automating legal reasoning gets a new round of interest every so often, and then everyone realizes the law is deliberately ambiguous and subject to interpretation, and that the legislators and judges like it that way.
I think the loss of professional autonomy is a bigger problem for the professions than automation is.