The Bacon Test

Over at The World's Fair, a challenge:

Anyway, this meme asks that you come up with your own scientific eponym. What's that exactly? Well, first read this excellent primer by Samuel Arbesman, which basically provides a step by step description of how to do this effectively. Then have a go at your own blog. If all goes well, I'd like to create a page at the Science Creative Quarterly, that collects (and links to) the good ones.

Of course there is already a formula for Best Bacon Butty, so I can't do anything about Bacon Butties, damnit. Well recently I've been thinking about how bad I am at calculations (probably because I'm teaching discrete math?), so here a la Alan Turing is my proposed eponym:

The Bacon Test

The Bacon Test is a proposal to test a human's capability to demonstrate raw computational power. The idea is to put a computer judge in contact with a human and a computer, and to see whether this judge, by connecting to these two subjects via a high speed data link and asking the participants to carry out fast computations, can distinguish between the human and the computer. If the judge cannot distinguish between the human and the computer, then the Bacon Test claims that the human is demonstrating computational power.

You may be interested to know that recent experimental evidence suggests that humans are not yet computationally powerful. I raced my computer at computing the partition function for an instance of a random bond two dimensional Ising model and lost. But I'm getting better, I swear! What's that about Moore's law and computers getting faster? Oh shoot.


More like this

In which I steal an analogy from Joe Emerson to explain the limits of quantum computing. ------------ As previously noted, a couple of weeks ago I went to Canada for the opening of the University of Waterloo's new Quantum Nano Center (their photo gallery includes one picture of me being interviewed…
In my discussion with Sal Cordova in this post, one point came up which I thought was interesting, and worth taking the time to flesh out as a separate post. It's about the distinction between a Turing equivalent computing system, and a Turing complete computation. It's true that in informal use,…
It sometimes seems like every day, some "intelligent design" bozo comes out with another book rehashing the same-old crap. I usually ignore it. But this time, I felt like the promotional materials for one of the new books really stepped right into my part of the world, rhetorically speaking, and…
As several [other][panda] [folks][pz] have mentioned, George Gilder has written a [new anti-evolution article][gilder-article] which was published in the National Review. [panda]: [pz]:…

When I first started programming computers back in '76, you threw a deck of cards at the machine and went off to have a sundae while the computer did your bidding. If the computer was slow, that was no big deal because you weren't waiting for it -- you were free to do what you wished. If you used too much time it just meant that they would put at a lower priority next time, or you could run out of computer time. In either case this meant that next time you could read the newspaper while enjoying your sundae.

Now computers are oodles of times faster, but I have to tend to it as if it were a newborn baby. When I install software I have to wait attentively and interminably for the computer to ask me to do crucial tasks like "clicking next".

If I meet Bill Gates I know what to ask. I don't want any money, I just want to know, what is it about "clicking next" that only humans are capable of this sort of calculation. Surely there is some genius out there that could write a program that would do this for me. The next Microsoft.

And if I try to go get a sundae, when I get back I find that the computer has powered itself off and left itself in a state that is incompatible with doing whatever I left it to do. To keep really long physics calculations to conclude, for example, applying Maxima to converting Painleve coordinates to Newtonian force equations, I have to stick around and occasionally move the mouse to prevent the computer from falling asleep while working for me.

As a young man, I looked forward to computers becoming intelligent and easy to use, but it seems that before this came about they first had to become neurotic and sickly. (Don't get me started on virsuses.)

Humans don't seem to be to good at fault tolerant computation. If we don't have paper and a pencil (essentially a noiseless ancillary system) then we become exponentially likely to fail to get the correct result.

Or at least I do.

I've had more sonnets published than any computer that I know. And yet, I've had poems published that were generated by a FORTRAN IV program that I ran on an IBM 1130 in 1966 (41 years ago). I keep waiting for someone to psychoanalyze the poet, at which point my trap will have been sprung. Turing evades the Poetry issue, to the annoyance of the Bacon-Post Caltech English B.S. brigade:

The Trouble with the Turing Test
Mark Halpern

... Consider Turing's suggested line of questioning with that strategy in mind:

Q: Please write me a sonnet on the subject of the Forth Bridge.

A: Count me out on this one. I never could write poetry.

Q: Add 34957 to 70764.

A: (Pause about 30 seconds and then give as answer) 105621.

Q: Do you play chess?

A: Yes.

Q: [describes an endgame position, then asks] What do you play?

A: (After a pause of 15 seconds) R-R8 mate.

"The first of these questions has no value as a discriminator, since the vast majority of humans would be as unable as a computer to produce a sonnet on short notice, if ever. Turing has the computer plead not just an inability to write a sonnet on an assigned subject, but an inability to write a poem of any kind on any subject. A few follow-up questions on this point might well have been revealing, even decisive for Test purposes. But Turing's imaginary interrogator never follows up on an interesting answer, switching instead to another topic altogether....