Can capitalism survive the artificial intelligence era? - Liberty.me

Can capitalism survive the artificial intelligence era?

Tags: AI artificial intelligence Capitalism

Andrew Lepkowski June 3, 2016 at 11:57 am

I have serious doubts that capitalism can survive the advent of strong artificial intelligence. It is terrifying to think how marginalized the individual will be when mass producible, strong AI dominates the labor market.

The resulting mass unemployment would lead to desperation, homelessness, riots, widespread crime, theft, murder, and chaos. The people, sadly, would clamor for more government; and the government would leverage the enormity of the situation to create “convincing” anti-capitalism propaganda.

You must be logged in to reply to this topic.

dL 1337 June 3, 2016 at 12:40 pm

A good video that addresses the topic.

One weakness of the analysis in the video is that it is too linear, that is it lacks feedback loops. As employable humans –> 0, who would be doing the consuming to drive the AI economy? There is this specter of this mass AI producing economy, but the question would be: for whom? Another thing to note: AI != no scarcity. AI robots would be powered by the same energy source that was around at the time of the horse and buggy. AI is not a sufficient condition for no scarcity.

Lastly, another thing the video missed, the sectors of human employment least susceptible to AI would be politicians,lobbyists, etc…the thing we libertarians would call the “looter class.” AI would drive employable humans not down to 0, but rather down to the looter class.

You must be logged in to reply to this topic.
You must be logged in to reply to this topic.

Andrew Lepkowski June 3, 2016 at 1:20 pm

I’ve watched that video a couple times about 2-3 years ago. It’s a pretty good intro to the topic.

As far as consumers go, I think that there would be some type of basic income or negative income tax (such as Milton Friedman’s concept) long before the advent of strong AI. By the time strong AI is broadly available, cheap and robust narrow-A.I./automation would long have since replaced a massive percentage of basic jobs and human tasks. Even with a negative income tax or basic income, there would likely still be mass poverty due to high inflation or even hyperinflation. The ratio of taxes collected to new money printed would be so enormously skewed that the currency would weaken rapidly.

The looter class, politicians, lobbyists, etc. would likely create laws and regulations to protect their “thrones” from would be A.I. competitors. Many citizens would eventually want to see the A.I. in charge of, some of (or possibly all of) the high-level judicial, legislative, and executive jobs. The govt./legislature would block this possibility long in advance (assuming the constitution doesn’t already make it impossible for many/most political seats to be “occupied” by A.I.)

You must be logged in to reply to this topic.
You must be logged in to reply to this topic.

Steve Wire June 3, 2016 at 2:28 pm

Strong AI will not mean disobedient AI. Strong AI will still be tools that will multiply worker productivity. I would be happy to have a staff of thousands of AI bots, boosting what I can accomplish enormously.

The assumption you have made is that boosting worker productivity will reduce the demand for workers. If that were true, we could have seen it before with any new technology. As the printing press, steam engines, electricity, automobiles, computers, etc. were each in turn invented, we would have seen a corresponding collapse in demand for labor. Why haven’t we? This is a fundamental question in economics.

Further: The population of the US has doubled in the last century, yet unemployment is still in single digits. Where did the jobs for 100 million new people come from? How can we understand the demand for labor better?

You must be logged in to reply to this topic.
- You must be logged in to reply to this topic.
  
  Andrew Lepkowski June 3, 2016 at 2:47 pm
  
  You should probably watch the video…this is not a luddite argument
  
  Couldn’t have said it better myself. Strong AI is by definition capable of human or beyond human intelligence. Workers won’t be needed anymore. Furthermore, there will be major economic problems long before AI even reaches that level of capability.
  
  You must be logged in to reply to this topic.
- - You must be logged in to reply to this topic.
dL 1337 June 3, 2016 at 2:42 pm

You should probably watch the video…this is not a luddite argument

You must be logged in to reply to this topic.
You must be logged in to reply to this topic.

Randall Chester Saunders June 3, 2016 at 3:42 pm

There are a number of things wrong with this new case for paranoia.

The first is that no one can predict the future, ever; and every new thing from from the certainty of world starvation to global climate change are just never true. Some of what is being predicted about the terribly misnamed technology, called AI will probably happen, but the entire scenario is just fiction.

I remember when the world was assured that computers were going to eliminate paper work. Have you been to a doctor’s office lately? …or any government office? …or any business office.

So, once all these marvelous AI robots are doing all these wonderful jobs, nobody is going to own them? Are the robots going to do all these things on their own? Who supplies the power for them? Who decides what jobs they are to do? Who designs and builds them in the first place? What if no one likes the products or services the robots provide and everyone just ignores them?

As for all the people who will, “be out of work,” because they didn’t bother to learn and keep up with technological changes, the real world is a tough world. Nothing stays the same. Everything changes.

I notice that all these scary scenarios of the future only worry about people who will be out of a “job,” not out of work, because a job is always something supplied by someone else. There is no guarantee of a job in life, but there is always opportunity to do productive work.

One scenario no one is mentioning is if robots really become as useful as is being described, they will become less expensive (as computers did) and will become available to individuals to use to produce…well just anything a computer can produce. Perhaps what robots will eventually eliminate is the large corporations, allowing every individual using their PR (personal robot) to be a self-employed entrepreneur.

Why is everyone so damn afraid to embrace the future?

You must be logged in to reply to this topic.
- You must be logged in to reply to this topic.
  
  dL 1337 June 3, 2016 at 8:03 pm
  
  Why is everyone so damn afraid to embrace the future?
  
  That’s not the way I would put it. The way I would put it is “not so quick to embrace tech triumphalism and evangelism.” I used to be a tech optimist. There was a time when I thought personal computing devices and the internet would liberate the individual from the state. The Crypto Libertarian “Declaration of the Independence of Cyberspace.” Well, It hasn’t turned out that way. The asymmetric empowerment that once favored the individual has now have turned on the individual. As it turns out, “ubiquitous internet, devices ,cheap storage + States + liberal democratic property rights” makes for an awesome surveillance state. The problem is that the Internet is made up of devices and connections that exist within national borders. As Julian Assange has noted in work, Cypherpunks: Freedom and Future of the Internet”
  
  The new world of the internet, abstracted from the old world of brute atoms, longed for independence. But states and their friends moved to control our new world — by controlling its physical underpinnings. The state, like an army around an oil well, or a customs agent extracting bribes at the border, would soon learn to leverage its control of physical space to gain control over our platonic realm. It would prevent the independence we had dreamed of, and then, squatting on fiber optic lines and around satellite ground stations, it would go on to mass intercept the information flow of our new world — it’s very essence even as every human, economic, and political relationship embraced it. The state would leech into the veins and arteries of our new societies, gobbling up every relationship expressed or communicated, every web page read, every message sent and every thought googled, and then store this knowledge, billions of interceptions a day, undreamed of power, in vast top secret warehouses, forever. It would go on to mine and mine again this treasure, the collective private intellectual output of humanity, with ever more sophisticated search and pattern finding algorithms, enriching the treasure and maximizing the power imbalance between interceptors and the world of interceptees. And then the state would reflect what it had learned back into the physical world, to start wars, to target drones, to manipulate UN committees and trade deals, and to do favors for its vast connected network of industries, insiders and cronies.
  
  Did it have to be this way? Does it have to be this way? I don’t know, but there is abetter chance of better outcome going forward if your eyes are open.
  
  You must be logged in to reply to this topic.
- - You must be logged in to reply to this topic.
Steve Wire June 3, 2016 at 4:14 pm

Okay, I’m watching the video.

Five minutes in, I’m thinking that a better example of AI than Baxter would be actual humans. What can we learn about human-like AI, like Baxter and beyond, from studying immigration’s impact on labor? I would love an answer to my earlier question: where did the most recent 100 million new jobs come from? And what is your opinion about open borders? I don’t see how someone could support open borders but fear strong AI.

The video’s answer to the above is to compare humans to horses. Wow. That’s very Malthusian. Obviously, comparing humans to animals is not insightful. Humans will surprise you if you do, and acting on your conclusions may raise ethical concerns. But it is suggestive about the kind of fear that strong AI invokes (in other people, not me).

Thanks for the discussion. I’m continuing from 5:00…

You must be logged in to reply to this topic.
- You must be logged in to reply to this topic.
  
  Andrew Lepkowski June 3, 2016 at 5:15 pm
  
  What can we learn about human-like AI, like Baxter and beyond, from studying immigration’s impact on labor?
  
  Not a whole lot. The language barrier that affects most immigrants (most are not fluent in English) slows and/or prevents most of them from acquiring many high-skilled jobs. Siri and IBM Watson already understand English better than the vast majority of new immigrants do.
  
  On a separate note, the cost structure of a human employee (immigrant or otherwise) is completely different from a robot or software bot. Bots have a high up-front cost (falling rapidly though) and extremely dirt cheap maintenance cost (and falling). The up-front cost to hire a person is low compared to a bot (for now), but their career accumulated cost is gigantic compared to a bot. Not to mention, personnel/labor costs are rising rapidly due to the “Affordable… Care Act” and ever-increasing regulations while bots do not suffer these cost influences. Also, people behave unpredictably (quitting unexpectedly, get sick, oversleep, lose motivation, etc.), which results in costs and losses that are very difficult to calculate and can range from inconsequential to the collapse of a business. Generally for bots, the potential consequences of a glitch or malfunction are well understood (per safety regulations) before the bot can be installed or booted up. They are far more predictable.
  
  where did the most recent 100 million new jobs come from?
  
  That’s a whole separate discussion. I never made any argument that jobs aren’t being created in the present. 100 million is an arbitrary number. You can pick any two distant points on a timeline and boldly proclaim huge numbers of jobs created.
  
  And what is your opinion about open borders?
  
  Open borders and a welfare state are incompatible. If you want to have open borders, you would have to completely minimize and completely scrap the existing welfare state. The second problem with open immigration is that roughly 70-80% of Latin Americans vote left-wing. IF that ratio remains unchanged, open immigration would inevitably transform the US into a Latin American style socialist country, gradually over time.
  
  You must be logged in to reply to this topic.
- You must be logged in to reply to this topic.
  
  dL 1337 June 3, 2016 at 7:15 pm
  
  Well, as I mentioned in original comment, I had a few problems with the analysis…too linear…there are complex feedbacks that have to be considered that are not easy to predict. So I would classify it as an open question, but one not subject to pat answers typical for technology displacement. Note: there could be a very good outcomes from it. For example, AI + cheap,storable renewable energy==no scarcity, so no one needs to work
  
  You must be logged in to reply to this topic.
  - You must be logged in to reply to this topic.
    
    Andrew Lepkowski June 3, 2016 at 8:22 pm
    
    The possibility of AI being able to autonomously produce an abundance of food, energy, medicine, etc. could have very positive impacts. However, with an extremely high percentage of common people having incomes (genuine income vs handouts) at or near zero, how many would be able to afford to buy strong AI machines to produce resources/wealth for them? Perhaps large groups of people would pool their meager savings and co-purchase/co-own an AI to produce income/resources for them. It would give a whole new meaning to income investing.
    
    You must be logged in to reply to this topic.
    - Dave Burns June 3, 2016 at 9:14 pm
      
      How many can afford a laptop?
      
      You must be logged in to reply to this topic.
    - dL 1337 June 3, 2016 at 10:37 pm
      
      obviously, no scarcity implies no market economy because there is no need of market exchange. Bots themselves are not sufficient for a no scarcity condition. They would have to be combined with something like very efficient solar battery storage. Essentially, then they would be powered by the sun. AI bots all the way down to the nanotech level…all powered by efficiently stored and converted renewable energy. “Need Not Reply” gets replaced by “No Need To Apply”
      
      That’s one possible outcome. I can think other less sanguine outcomes…
      
      You must be logged in to reply to this topic.
    - Andrew Lepkowski June 3, 2016 at 11:33 pm
      
      Things are going to get very weird. That’s one thing that we can count on. This current era will seem very alien to people in the not too distant future. I think that “no scarcity” is a misleading way of describing an abundance scenario. Certain things like gold, antiques, rare gems, jewelry, genuine artwork, prime real estate locations, rare luxury items, etc. will still be scarce more or less. Trade will still exist for anything that is not hyper-abundant.
      
      You must be logged in to reply to this topic.
- - - - You must be logged in to reply to this topic.
Dave Burns June 7, 2016 at 4:41 pm

Ask the AI to solve the problem.

You must be logged in to reply to this topic.
You must be logged in to reply to this topic.

Randall Chester Saunders June 7, 2016 at 7:45 pm

“Ask the AI to solve the problem.”

Brilliant! Dave.

If AI is a problem, and I doubt it is, it should certainly be able to solve the problem.

You must be logged in to reply to this topic.
- You must be logged in to reply to this topic.
  
  Dave Burns June 11, 2016 at 4:36 pm
  
  Should I have put in an emoji? Didn’t want to ruin it.
  
  You must be logged in to reply to this topic.
- - You must be logged in to reply to this topic.
Frank Maruschak June 16, 2016 at 5:46 am

Laziness, not necessity, is the mother of invention! If one observes that there are individuals with a strong propensity for CONTROLLING the lives & actions of others, and IF such control is considered a CRIME against society, then AI could be a problem. The more the AI is concentrated in the hands of such individuals the greater the problem, especially when the individuals are “leaders”. If there is someone with the resources for the exposure of how AI will be used to effect the up coming election NOW will be the time to speak UP. IT is NOT the AI, but the USE of the AI that will make the difference.

You must be logged in to reply to this topic.
You must be logged in to reply to this topic.

You must be logged in to reply to this topic.