3 ms·
I gave the Crossword puzzle to Claude and got a correct response[1]. The fact that they are comparing this to gpt4o and not to gpt4 suggests that it is less imp
by csomar 2y ago
I gave the Crossword puzzle to Claude and got a correct response[1]. The fact that they are comparing this to gpt4o and not to gpt4 suggests that it is less impressive than they are trying to pretend.
[1]:
Based on the given clues, here's the solved crossword puzzle:
+---+---+---+---+---+---+
| E | S | C | A | P | E |
+---+---+---+---+---+---+
| S | E | A | L | E | R |
+---+---+---+---+---+---+
| T | E | R | E | S | A |
+---+---+---+---+---+---+
| A | D | E | P | T | S |
+---+---+---+---+---+---+
| T | E | P | E | E | E |
+---+---+---+---+---+---+
| E | R | R | O | R | S |
+---+---+---+---+---+---+
Across:
ESCAPE (Evade)
SEALER (One to close envelopes)
TERESA (Mother Teresa)
ADEPTS (Initiated people)
TEPEE (Native American tent)
ERRORS (Mistakes)
Down:
ESTATE (Estate car - Station wagon)
SEEDER (Automatic planting machine)
CAREER (Profession)
ALEPPO (Syrian and Turkish pepper variety)
PESTER (Annoy)
ERASES (Deletes)
- thomasahle 2y agoAs good as Claude has gotten recently in reasoning, they are likely using RL behind the scenes too. Supposedly, o1/strawberry was initially created as an engine for high-quality synthetic reasoning data for the new model generation. I wonder if Anthropic could release their generator as a usable model too.
- deisteve 2y agowhile i was initially excited now im having second thoughts after seeing the experiments run by people in the comments here on X I see a totally different energy more about hyping it on HN I see reserved and collected take which I trust more. I do wonder why they chose gpt4o which I never bother to use for coding. Claude is still king and looks like I won't have to subscribe to ChatGPT Plus seeing it fail on some of the important experiments run by folks on HN If anything these type of releases that air more on the side of hype given OpenAI's track record
- valval 2y agoI think people are wrong just about as often here as anywhere else on the internet, but with more confidence. Averaging HN comments would just produce outputs similar to rudimentary LLMs with a bit snobbier of a tone, I imagine.