speculating on Colossus
I would like to spend some time today weaving together some old treads. I’ve written a couple of times before about Colossus, here (in 2018) and here (just last year). I’m not entirely sure why I keep coming back to this old story, written long before we had any strong evidence of just how computers will change our lives.
Perhaps it’s because I find it such a powerful cautionary tale. We build this machine, capable of gathering all available information and being able then to make binding decisions in the world based on its understanding of that information. And then it goes awry. This is an old story, the creation running amok. It is a reflection of that old conundrum, that being able to do something is not in itself a reason for then doing it. Can does not equal should.
So the question came to me recently: who is building Colossus. We are building more and more powerful tools; we see this in the news every day. We encounter them in our daily lives. And we don’t know how they work. This observation is fleshed out in an essay by Stephen Wolfram, and this observation is part of this ecosystem of ideas, these tools we build that we may not understand how to use.
But I’m sure that someone out there, some government perhaps, is building a Colossus. A computer, a programme, through which all information flows. A computer, a programme, that we can then ask, what are others doing. The quality of the answers will depend on the quality of the information, which is why I think it will be a government. Governments are the keepers of secrets; the image that comes to mind is the warehouse at the end of the Raiders of the Lost Ark.
If there were a machine that knew everything that was known, that possessed in its internal structures every fact known, what question would I ask of it? In Jokester, the question is the origin of humor. Others have their questions, but what would mine be, and I’m not sure.
What’s interesting to speculate about gets back to my day job. If this machine knows everything, then it knows all of the math we know, all of the math we’ve discovered (and yes, discovered and not created, and that is a different conversation entirely), and I suspect my first question would be a math question. Do we know enough to answer the questions that I’ve carried for years? Decades? For some, almost certainly not, but there are others, questions that lurk at the outer edges of the math I know well, and perhaps these.
And so, speculation. I’ll continue to ponder this question of which question, and speculate as well on what sort of answer we might expect to get.
multijimbo 