3 ms·
> My guess is that image understanding will improve. The "it will get better" assertion fails on what's known as the "first step fallacy". The analogy I like
by cratermoon 2y ago
> My guess is that image understanding will improve.
The "it will get better" assertion fails on what's known as the "first step fallacy".
The analogy I like to use is the idea of a ladder to the moon.
We can build tall self-supporting ladders,
and ladder technology has advanced considerably since they were first used at least 10,000 years ago.
More recently, in 1862,
John H. Balsley made the step ladder safer by changing the rounded rungs to flat steps.
Henry Quackenbush patented the extension ladder in 1867.
Ladder technology continues to progress,
will we someday build a ladder tall enough to climb to the moon?
Do you see the fallacy?
https://thebullshitmachines.com/lesson-16-the-first-step-fallacy/index.html https://thebullshitmachines.com/lesson-16-the-first-step-fal...
- someothherguyy 2y agoNo one is attempting to build ladders to the moon though?
- namaria 2y agoNo, but some people claim that compressing text into vector spaces and using the patterns to extend bits of text will somehow do more than extend texts in a sometimes plausible, sometimes nonsensical way if we just: add enough data, and process it for long enough.
- someothherguyy 2y agoNot saying its a non-sequitur, but both arguments seem to follow a false premise.
- segh 2y agoThis is the crux of the issue. Whether you think this is like extending a ladder to the moon, or more like we figured out how to get to the moon and are now aiming at Jupiter.
- skybrian 2y agoYes, I've heard that analogy before, but it's more of a question than an answer: is there an unbridgeable gap that prevents continuous improvement, or not? Researchers will have to figure out whether such a gap exists. Since I'm speculating, not making a claim, there's no fallacy here.