October 29, 2009

Artificial General Intelligence, Automation and Economics

J Storrs Hall points out employment test for Artificial General Intelligence that was suggested by Nils Nilsson

AI programs must be able to perform the jobs ordinarily performed by humans. Progress toward human-level AI could then be measured by the fraction of these jobs that can be acceptably performed by machines. The list of jobs are specifically listed from America's job bank. Receptionist, paralegal, Meeting and Convention Planner etc...

There are automation systems, internet and cellphone applications and business processes now that allow people to not use the services of staff to perform many of the listed tasks.

Evite, and Calendar applications and unconferences can remove the need for meeting and convention planners or reduce staffing levels. Enabling self-organization using computer applications and self organizing processes "automates" and reduces staffing. Effort and staff then can go to specific recruiting and booking of keynote speakers or other less automatible maximum value add tasks.

The need for paralegals can be reduced with streamlined processes and reduced bureaucracy.

I had made this case in a previous article that we have experienced and will continue to experience waves of process changes (some combined with some form of automation). [complex and more capable vending machines, self-service retail checkout, etc...]

Hanson v Ford Debate on Implications of Intelligent Machines

(H/T to Martin Ford and to Michael Anissimov at Accelerating Future and Michael also commented on the J Storrs article)

Martin Ford tries to make the case that

Economic growth at any level—let alone of an order of magnitude beyond what we are accustomed to—is fundamentally incompatible with wages that are falling dramatically for the vast majority of workers. We might, perhaps, have vigorous economic growth if the falling wages applied to only a minority of human workers, but it is very difficult for me to conceive of a way in which such growth would be compatible with wages falling across the board—or even for the bulk of workers.

My reaction: There was decades of stronger economic growth for the US economy as a large percentage of jobs had reduced wages as manufacturing and other work was outsourced to cheaper workers overseas and to lower wage illegal immigrants. There have been transitions in large industries where jobs were massively reduced because of automation or new processes. The shift out of agricultural jobs and more automation in manufacturing. Reductions in HTML coders and programmers with the shift to Web 2.0.

Robin Hanson replied to Martin Ford:

The standard views of techies about what techs will be feasible might be wrong, and the standard views of economists of how to forecast tech consequences might be wrong. And it is fine for contrarians to try to persuade specialists they are in error, though contrarians would be wise to at least understand the standard view before trying to overturn it. But surely what the world needs first and foremost is to see and take seriously the simple combination of the standard views on such important topics.

Yes, a sudden unanticipated change would be disruptive, but no disruption does not imply falling production. Ford needs to learn some economics, or listen to some economists.

Martin Ford replies with a repeat of his claims of economic doom from automation.

I have equated the advance of better artificial intelligence and robotics as being comparable to wave after wave of outsourcing.

Someone or group will own or control or enable each wave of artificial intelligence and robotics. Just as there are those who primarily profit from different versions of the Linux operating system or the MySQL database. The value and profit is from being the go to place for related services or from continuing to improve the system or method. The open source model is a way to look at AGI/robotics in that the programmers who are contributing to building the capabilities and value of MySQL and Linux are probably smarter and more productive than say the executives and owners of Redhat but the "dumber" people who own the right things are in the right legal position are the ones who profit.

There will systems without consciousness that can perform certain task sets perfectly. There are systems now that can play a perfect game of checkers. Any person could compete against anything in the context of a game of checkers and play perfectly using such a system. A person with that assistance could not be outcompeted in that context. How much economy can there be with good enough human provided goods and services ?

Adrian Slywotzky has written several books about finding ways to migrate your business to a profitable area when your old business and old business models are challenged.

Enabling more entrepreneurship and getting people with the ability to add value in different ways will enable more success. If the AGI (artificial general intelligence) and robotic/automation end game is that people do not have "jobs" then people need to own the broad based mutual funds which have a piece of the companies that are profiting and winning. Also, as the final automation waves are happening then people should be looking ahead and being prepared to retire. Have the money and assets (houses, land, devices) and ownership of sustainable revenue streams. Even if people were cut out of the "robot/AGI economy" they can still trade amongst themselves. The Amish have a functional and sustainable system and choose not to do certain things. Even if there is another level of robot/AGI interaction that would not detract from what individuals have and can do. The majority of people now are not really involved in high finance in New York or Tokyo or the other financial centers. How much of the current world economy are most people connected to in any way ? This would be even more the case for developing countries in Africa. If subsistence and health issues are taken care of then how much does it really matter if there other levels to an economy ? If Warren Buffet is a multi-billionaire or interplanetary trillionaire who is enabled with AGI/robotics, so what ?

blog comments powered by Disqus