Sunday, August 7, 2016

Will neural lace save us from the super-intelligent robots?

If neural lace sounds like something out of a science fiction novel that is probably because that is where the idea originated. A couple of months ago, however, a group of scientists published a paper about injecting an ultra-fine mesh into brains to create neural lace. The mesh has been tested on mice, which survived the implantation and are thriving. Suggested uses for neural lace include “monitoring brain activity, delivering treatment for degenerative disorders like Parkinson’s, and even enhancing brain capabilities”.

In terms of economics, innovations to enhance brain capabilities seem to have many characteristics in common with other innovations. It seems reasonable to expect that, as with other successful technological advances, entrepreneurs developing the technology will initially be able to charge a high price for it. As suggested by Anders Sandberg, of the Future of Technology Institute at Oxford, the price of some brain enhancements (pills and gadgets) can be expected to fall as a result of competition, while the price of service-heavy enhancements (genetic engineering) is more likely to stay expensive. On that basis, there could be potential for the price of neural lace to fall fairly rapidly and for it to become widely used within a few decades, if it offers substantial benefits to users.

 Another reason to expect the price of neural lace to fall rapidly is that research in this area is likely to be undertaken in several different countries. It seems unlikely that the government of any one country will be able to protect the rents of entrepreneurs who develop the technology by suppressing international competition for a prolonged period.

Elon Musk’s proposal to develop neural lace to prevent humans becoming house pets of intelligent machines provides a further reason to expect neural lace to be priced for the mass market. Elon has several other highly ambitious projects on his plate, but he seems willing to add neural lace to the menu. His attitude: “Somebody’s got to do it. If somebody doesn’t do it, then I think I should do it”.

Elon Musk established OpenAI as a not-for-profit venture, so his neural lace project would presumably not be aimed at maximizing shareholder value. Perhaps Elon can attract sufficient potential investors (or donors) to be able to fund the necessary research and development.

So, why not join Elon’s fan club and help him begin work on neural lace as soon as possible to save our descendants from the robots? Michael Cook, a bioethicist, has given three good reasons to proceed cautiously:
“Take privacy. If you can hack a computer, you can hack a brain. Integrating your memories and cognitive activity with the internet allows other people to see what you are doing and thinking 24/7 — a kind of upscale parole bracelet.
Take autonomy. In our culture, this is the most cherished of our personal values. But once brains are integrated into an information network, they can be manipulated in increasingly sophisticated ways. And since technology always serves its owner, we could easily become the tools of Google or the government.
Take responsibility. There might be no crime, as the neural lace could shut down the ‘hardware’ whenever passions threaten to overwhelm social norms – as defined by the network.”

Elon Musk’s claim that humans are already cyborgs should be rejected. The fact that we, as individuals, may have a presence on social media does not mean that we have ceased to be uniquely human. Individuals can have brain implants and still be uniquely human.

Neural lace will not be worth having unless it can be developed in such a way as to enable humans to protect the privacy, autonomy and responsibility that is integral to their individual flourishing.  


Leah said...

I was going to make the same argument until I came to the highlighted quote by the bioethicist; anything digital can be hacked. What is more, is it really a foregone conclusion that the machines are going to take over?

Winton Bates said...

Hi Leah, it is not a foregone conclusion that the machines will take over. I don't have technical competence in this area, but it is apparently likely that artificial general intelligence will advance to the point where it exceeds human intelligence sometime this century. What happens after that depends on what objectives the machines have been programmed to pursue. Nick Bostrom's book, which I discussed here in June suggests that it might be difficult to ensure that the intelligent machines are benign.

I am not sure what we should do to protect future generations. It would be stupid to try to prevent future advances in AI. Apart from the potential loss in human well-being from benign use of the technology, regulation to prevent research is unlikely to be successful on a global scale. If the research is undertaken in several different places that may provide some protection against a single machine running amok. Research by defence departments may provide some protection against AI use in cyber hacking, but the efforts of governments to get an edge over potential enemies Is also potentially part of the problem. Perhaps our best defence lies in further development of protocols for protection of individual privacy and autonomy.

Leah G said...

I have always though that technology was a double edged sword - it can be used for good or ill. I found this article very thought provoking:

Winton Bates said...

Thanks Leah. That article certainly makes points worth thinking about. My initial thoughts are that technology does have a bias, but the main impact is to expand the possibilities available to us. Many people have learned how to turn off the TV but it took some time. (It took me about 40 years.) Perhaps we can still be optimistic about new technology if we can be optimistic about human nature and institutions. I must try to write something about this over the next few weeks. I should take another look at Virginia Postrel's book about the future and its enemies before I do that.

Leah G said...

I agree with you Winton. It seems to me that technology is an extension of us, in the same way that the market and other human creations are extensions of us. Human actions are what matters.

Some people try to make the argument that market is fundamentally immoral. But it is prima facie neutral or amoral. A voluntary, mutually agreed upon exchange, free from coercion, which benefits both parties, can hardly be immoral. Where as fraud, cronyism, etc. these are examples of immorality in the marketplace.

Winton Bates said...

Yes Leah, some economists have been taking an interest in the impact on economic development of cultural norms and levels of interpersonal trust. Low levels of trust result in high transactions costs.

At the same time, I think there is some validity in fears of markets being extended to aspects of human interaction where "commodification" is not appropriate.