Display MoreI agree with Arthur Clarke on this issue. He said something like, "I am sure that conscious, sentient machines are possible because I carry one on my shoulders." In other words, if biological neuron-based brains that evolved on earth can be conscious, we have no reason to think that similar brains based on other biology elsewhere in the universe are also conscious, or that technology-based machines with similar neurons cannot also be conscious. To say it is impossible is to imply there is something magical or ineffable about brain tissue that cannot possibly be replicated in other media. When I say "replicate" I mean something similar to the way electroactive polymers replicate muscles, or ventricular assist devices (VAD) replicate the heart. I mean they perform the same function, with the same outcome, even though the underlying physical properties are different.
The "ineffable" hypothesis -- that there is something science can never understand -- has been applied to many phenomena in the past, especially the origin of life, genetics, and the ability of one cell to reproduce an entire complex organism. These things were once considered miraculous and forever beyond our grasp. Nowadays we know exactly how they work. The basic principles are taught in junior high school.
To address the author's points:
"a) We do not understand the neural basis of consciousness in humans, nor do we have a clear, uncontroversial philosophical answer to the hard problem of consciousness . . ."
We don't have to understand it completely. No one knows exactly, in every detail, how combustion works. We learn more about it every year. Despite our incomplete knowledge we have been using fire successfully for 400,000 years. We do not know every aspect of E. Coli biology, but we know enough to control infections and use E. Coli to manufacture many useful things such as human insulin.
"This makes a theoretical, top down approach to building machine consciousness difficult."
So what? A bottoms up approach has worked for nearly every technology we use. We had no top down approach to fire, metallurgy, stone cutting or anything else before 1650.
"b) We don’t know if building conscious machines is even compatible with the laws of nature."
Of course we know this! There are billions of conscious machines all around us.
"c) We don’t know if Big Tech will want to build conscious machines due to the ethical sensitivity of creating conscious systems."
That is a completely separate issue, having nothing to do with the technical issues. If Big Tech in the next 50 years will not do this, perhaps it will 100 years from now, or 300 years from now. This is like saying big automobile manufacturers did not want to make electric cars in 1980. That is true, but it did not mean electric cars were not feasible or practical in 1980. Manufacturers made a business decision not to develop electric cars, which turned out to be a mistake.
"d) We don’t know if building conscious AI would be technologically feasible — it might be ridiculously expensive."
Of course it is not technically feasible today. This is like saying in 1960 that "a desktop computer than does 3 billion operations per second is not technically feasible." Everyone knew that back in 1962. It would have been ridiculously expensive. Such machines are everywhere today, and they cost practically nothing. There is no reason to think that a silicon based intelligent machine will take large amounts of materials, or large amounts of electricity, or expensive materials, or processing and manufacturing more expensive than today's computers. Obviously it will take techniques we have not yet discovered, but once you invent a technique it costs nothing after that. Once you teach a robot how to balance, deal with the center of mass, and dance, robots will be able to do that for the rest of history at no additional cost. (See the video below.)
There is no reason to think that an intelligent machine with approximately the power of a human brain will be much larger than a human brain. Unless we happen to want one that is a million times more powerful than a human brain, which might be a handy thing to have. Sometimes it is useful to have a machine that is far more powerful than any human or biological equivalent, such as an airplane that flies faster than any bird, or a supercomputer that does more computations per second than the entire human race could with pencil and paper.
Granted, some biological data processing devices are far better than human technology. DNA storage is so compact, all of the world's data could fit into a liter of DNA. It it longer lasting. DNA kept in proper conditions will last for hundreds of thousands of years. DNA is far ahead of human data storage today . . . but people are working on DNA storage, and similar molecular data storage. I expect in 20 to 50 years our data storage will be as good as DNA. It might actually be DNA. Prof. Church at Harvard has stored and reproduced book in DNA. Storage and reading back is slow, but reproduction is far faster than any human technology. It is likely that small, cheap machines to read and write DNA will be developed, giving us hard disk-like storage billions of times denser than anything we now have. There is no reason to think this will not happen.
Dancing robots. This ability will soon be in all robots at no cost, like our ability to store gigabytes of data is in every computer, cell phone and camera, at practically no cost:
After trying out Chat GPT and see the reaction from insiders I think that the step to a get consciousness is just about creating a recursive machine e.g. the output is in some more or less clever way put to the input side and it will be able to simulate things an develop dynamically. We really got enough training material to do this after myself evaluating Chat Gpt. Currently this is not something that is done, but when people start playing with that for these powerful networks I'm sure that in the end it will develop conciusness. You are right that the goal is not to do this when we create tools today. But sooner or later the missing connection will be connected and I bet that in some server hall in some defense building real life can take it's first breath maybe even today. Hmm maybe that's' why I started to follow Arnold Schwarzenegger on facebook and not because my body is in sore need of training.