The Artificial Intelligentsia

The argument at the moment seems to be "define a problem that a computer can't do that a human brain can"... "I can't because expressing that problem is beyond the machinery I have developed for cognition, and it may always be".

What is certain is that there are uncomputable problems, but are any of the problems that humans solve in order to speak, act, socialise uncomputable? Some people think that because they are solved within the physical universe then they must be computable but that implies that the physical universe can be simulated (in principle) by a universal turing machine, since we can express problems (using the machinary of the universe) that a universal turing machine can't do then there is a gap that permits the possibility that there may be some process which is not simulatable by a universal turing machine.

In my belief free will/autonomy/initaitive/creativity are expressions of that process, but belief is not an argument.

The article you reference makes the same mistake as Searles Chinese room argument. It somehow assumes there is something magical about the human mind.

I had a brief argument with Robert Epstein (the author of that article) because I find the argument that humans don't actually store information to be quite misleadning and missing the point.

The most obvious mistake is this:

"We don’t store words or the rules that tell us how to manipulate them. We don’t create representations of visual stimuli, store them in a short-term memory buffer, and then transfer the representation into a long-term memory device. We don’t retrieve information or images or words from memory registers. Computers do all of these things, but organisms do not."

This is both true and false. It's true we don't do that like a computer but its wrong to claim that computers fundamentally do that too.

That happens several layers of abstraction up and so a computer fundamentally doesn't actually store an image or a word either it manipulates atoms and turns circuits on and off and several layers of abstraction up it gets translated into meaning first by machines then by humans.

Anyone who have a hard time believing machines can become sentient should first ask themselves why they have a harder time believing that than accepting that dumb immaterial matter somehow have become the pattern recognizing feedback loops that are us.

I believe there is an alternate reading you may be missing. Viewed in a different light the example of Go you cite is not that different from John Henry vs. the steam engine. The history of industrialization is the replacement of humans by machines for specific tasks. People tend to oversell current AI technology as somehow deviating from that long path. It's a fair point to make.

For me the interesting theme was the exploration of the character and dubious success of the mysterious Jim, who is using his connections to ride a wave of poorly understood and possibly malevolent technology to a grand house in the country and membership in the upper classes--almost like flotsam on the rising tide of economic progress. There's a lot of Gatsby in there and some hints of Graham Greene's quiet American as well. Just as you can argue that AI the technology is part and parcel of industrialization, its social effects recall recurring conflicts in American society and culture that authors like Fitzgerald have been exploring for 150 years.

As for other critiques of the style from the Hackerati I would just say yes, it seems like the work of a young writer. Good writing is hard to achieve and it's typically preceded by a lot of bad writing. To paraphrase Senator Palpatine, we will watch his career with great interest.

Fixed: typo/it really is hard to write well

I enjoyed the article as well (see my comments above). But I would debate your takeaway. The money quote is in the last sentence:

> Three months later, Predata secured a second round of venture capital funding.

People like Jim Shinn will always find a way. At least that's the argument the author seems to be making.

From all I've been able to see, that statement

"... judge the merit of a new idea in AI according to the perceived intelligence of its developers."

about information technology VCs and AI is just totally wrong: I don't believe VCs do that. Why? Generally, from 50,000 feet up, it's too far from the norms of the accounting, banking, and investing communities respected by the limited partners of the VCs. Uh, the limited partners (LPs) are where the VCs get nearly all the money they invest, and the limited partners are conservative people, managers of pension funds, university endowments, etc. Not only do the VCs not do that, the LPs won't let the VCs do that!

Instead, about the shortest believable view I can see is, VCs look for traction that is significant and growing rapidly in a market large enough to permit a company worth $1+ billion in a few years.

The VCs view of traction is a weakening of the usual measures the accounting, banking, and investing communities use and respect of audited revenue and earnings.

So, sure, the best form of traction will be earnings, then next best, revenue, next best lots of interested customers, e.g., advertisers willing to pay for eyeballs, then last best, just lots of eyeballs. In these norms, intelligence, brilliance, AI, technology, etc. are mostly publicity points, window dressing, the wrapping paper on a birthday gift, and with a dime won't cover a 10 cent cup of coffee.

In a sense, the VCs have a good point, more from insight into humans and the real world than anything in a pitch deck: (1) With technology, it's too easy to push totally meaningless, useless BS. (2) Carefully studying core, deep, difficult technology is just too darned difficult to be practical for the VCs.

Or the investors believe in a Markov assumption: The future of the business and the technology from the past are conditionally independent given the current traction, its rate of growth, and the size of the market. To be clear, this Markov assumption does not say that the technology and the future of the company are independent.

The stories in the OP about the company Predata, to abbreviate "predictions from data", are good: The company was floundering around with guesses about what would work, e.g., for predicting terrorist attacks, that were like something from smoking funny stuff.

But here is one big place the VCs and technology are going wrong: We do have some terrific examples of how to do well. The examples are from the past 70+ years of the unique world class, all-time, unchallenged grand champion of using advanced, even original, technology for important practical results -- the US DoD.

A grand example is GPS. GPS was by the USAF, but it was a refinement of an earlier system by the US Navy, for navigation for the missile firing submarines and started at the Johns Hopkins University Applied Physics Laboratory JHU/APL. At one time I worked in the group that did the original work and heard the stories. A key point: The original proposal was by some physicists and almost just on the back of an envelope. Soon the project was approved and pushed forward with a lot of effort. Then, presto, bingo, it all worked just as predicted on the back of the envelope. E.g., a test receiver on the roof navigated its position within one foot, plenty accurate enough for the US Navy.

So, net, for project selection and funding, here is the shocking, surprising, point that the VCs miss: Really, given the back of the envelope work, the rest was relatively routine and low risk.

And the past 70+ years of the US DoD is awash in comparable examples.

In blunt terms, the US DoD has a fantastically high batting average on far out projects evaluated just on paper. Given good evaluations of the work just on paper, the rest is relatively routine and low risk.

Well, that project funding technique does not fully solve the problem of the VCs: The VCs also need to know that the resulting product will have big success in the market. But for that there is an okay approach: The dream product would be one pill taken once, cheap, safe, effective, to cure any cancer. In that case, the technology is so good for such an important practical problem in such a large market that there's no question about making the $1+ billion. So, from this hypothetical lesson, net, need the technology to be the first good or a much better solution, a "must have", for a really pressing problem in a big market. So, right, this filter would reject Facebook, Snap, and more. So, right, need to start with a really big problem where with new technology, say, as in the US DoD examples, can get a "must have" solution for a really big problem, and Facebook and SNAP are not such problems. Just what are such problems? That's part of the challenge. But with current VCs, come up with such a problem and a solution on paper, with brilliant founders, with AI, etc., then still will need more than a time to cover a 10 cent cup of coffee. Again, to get VCs up on their hind legs, bring then good data on traction, significant and growing rapidly in a large market; if the secret sauce technology helps, fine; brilliant founders, fine; even if there is no technology, fine; in all cases, what really matters is the traction.