[Sigia-l] Findability

[Karl Fast]

> Imagine it's 2003....in 25 years it would be commonplace to search
> through a three billion document set in less than a second, return
> results that most people were satisfied with most of the time and a
> computer would understand them. Not perfect mind you, but good
> enough that...
> 
> If I said this you'd think I was bonkers. But you'd be wrong.


Touche.

However, there is a subtle difference between my argument and yours.
Mine is primarily about scalability and computation. People have a
hard time understanding exponential growth. I 1983 the idea that
a computer could process that much stuff that fast seemed pretty
crazy. If you ran the numbers it's not that surprising.

Now, consider your scenario in which in 25 years a computer could
understand the results in an intelligent way that approaches human
understanding (not perfect, but good enough). 

If that happens through continued scaling a'la Moore's law, it would
suggest that human beings are essentially computing machines and
that our brains are, essentially, really good pentium chips.

I kinda doubt that.

On the other hand, it could happen through some sort of radical
breakthrough. Maybe the Stephen Wolfram "new kind of science" thing.

That seems a more plausible scenario to me. 

But to say that we'll get there (to an IR system that truly
understands the documents in front of it) by doing everything we do
now, only bigger and faster...that does seem bonkers to me.



--karl