4 ms·
“Claude gets it” No. It does not. It does not understand anything. Stop anthropomorphizing bots!
by frizlab 10mo ago
“Claude gets it”
No. It does not. It does not understand anything. Stop anthropomorphizing bots!
- onion2k 10mo agoStop anthropomorphizing bots! They hate that.
- jiggawatts 10mo ago2025 is the year where unironically we have to mindful of the emotional state of our computer software, otherwise our tools may just flip the table in frustration and rm -rf the codebase.
- dboreham 10mo agoHow do you know whether a human brain understands something?
- shimman 10mo agoProbably the same way that I can be assured your interpretation of red is mine.
- kshahkshah 10mo agoSome people are colorblind. Some people have more or less cones and rods. Our interpretation of colors is certainly not the same
- mrbungie 10mo agoYou should steel-man the argument. GP is talking about qualia, obviously for the sake of the argument you assume the comparison is between two people with similar eyes.
- hyperhello 10mo agoSteel-man is such a weird expression. There are no steel men. How about saying "The opponent's best argument".
- collingreen 10mo agoThe steel men (armored enemy knights) are exactly the inverse of the straw man (training dummy) metaphor. I think it's a fantastic term since it directly addresses the point (tackle the best opposing arguments head on instead of a poor subset/facsimile of them), it fits within the existing straw man metaphor, it's terse, and it's very clear.
- hyperhello 10mo agoThanks for replying.
- nawgz 10mo agoThe wild success of traffic lights disagrees with your statement.
- nickthegreek 10mo agoThe wild success of traffic lights comes from having 3 colors at fixed positions. You put those 3 colors in a single color changing light and I would assume the accident rate would measurably increase.
- evilduck 10mo agoThe fact that a single emitter traffic light that simply varies its color doesn't exist also disagrees with your statement.
- inetknght 10mo agoThe wild success of traffic lights is only wildly successful to those who aren't color blind. Do some reading. Wikipedia: https://en.wikipedia.org/wiki/Color_blindness https://en.wikipedia.org/wiki/Color_blindness > The colors of traffic lights can be difficult for red–green color-blind people. This difficulty includes distinguishing red/amber lights from sodium street lamps, distinguishing green lights (closer to cyan) from white lights, and distinguishing red from amber lights, especially when there are no positional clues (see image). Publication from 1983: https://pmc.ncbi.nlm.nih.gov/articles/PMC1875309/ https://pmc.ncbi.nlm.nih.gov/articles/PMC1875309/ > All but one admitted to difficulties with traffic signals, one admitted to a previously undeclared accident due to his colour blindness, and all but one offered suggestions for improving signal recognition. Nearly all reported confusion with street and signal lights, and confusion between the red and amber signals was common.
- nawgz 10mo agoWhat a horrendous counter-argument. "People with notable perception issues don't perceive the same" is insanely obvious.
- onraglanroad 10mo ago
- einarfd 10mo agoColors and color names are culture dependent, and you are not guaranteed that people in different cultures agree on what color something is. The most famous of these discrepancies is Japan and green vs blue, or why does Jenkins by default use red, yellow and blue instead of red, yellow and green. So I would urge using something other than colors as an example of shared human experience.
- shimman 10mo agoYou're talking about culture and names, I am not. We all live in an objective reality where a red spectrum of light exists and we can differentiate among other colors. This is the contention with the person I am replying to, they're acting as if objective reality doesn't exist. Humans can think, LLMs cannot. If you can't admit to this there's nothing else worth discussing, but please don't mind my hands covering my wallet as I slowly back out of the room.
- shermantanktop 10mo ago> we can differentiate among other colors. Some of us color-deficient people can’t. I only accept that stop signs are “red” because all the normies say it is. Your point stands, but color perception is not the best example for it.
- Groxx 10mo agothank you for providing evidence that some do not.
- onraglanroad 10mo agoThat's weird. You aren't actually being sarcastic but literally believe the opposite of your post!
- greekrich92 10mo agoDenial of an objective reality is a symptom shared by various strands of bad thinking our current era. LLMs do not think. That's reality.
- immibis 10mo agoNobody knows what thinking is. A real human brain can be simulated with matrix multiplications.
- Eisenstein 10mo agoWhat could an AI do that would convince you that it is able to think?
- bawolff 10mo agoWin a (properly conducted) Turing test
- greekrich92 10mo agoAI or an LLM? LLMs are matrix multiplication word guessing machines, so nothing. AI doesn't exist yet.
- Eisenstein 10mo agoThat's an interesting difference to make. You are defining an AI based on a result (that it displays intelligence) yet dismissing another thing based on method (how it works). By that logic it would have to work by a method which you already define as intelligent in order to display intelligence. Isn't that circular? Let's remove AI or LLM from this question and ask if there is anything any non-biological system could do to convince you that it was thinking?
- rcxdude 10mo ago"think" is not sufficiently defined to assign an objective truth value to that statement.
- throw-12-16 10mo agoSave it for your blunt rotation.
- efilife 10mo agonot again
- willparks 10mo ago"Claude has been trained to handle this the right way"