Quick Thoughts
Region: US
Wednesday 28 January 2026 21:40:18 GMT
940599
35570
382
842
Music
Download
Comments
Tre’Jan Vinson :
How can I gamble on this ?
2026-01-29 03:56:23
4120
Diatride :
"Orange or yellow" please tell me you guys see orange and nothing close to yellow. Genuinely doing a colorblind check on myself
2026-01-29 06:29:30
1997
Longue-Vue :
fyi that's definitely blue and orange
2026-01-29 01:05:17
1825
ccccccccory :
this used 40 gallons of water and enough power to run a moderate house for half a day.
2026-01-29 01:50:05
769
winston_peters_1945 :
You could have just chucked this in excel dawg
2026-01-29 23:58:18
752
Just us :
Have you checked the work of any of their math? I’ve found AI bad at real math but good at saying thing that sound right.
2026-01-29 04:08:14
306
Ellen :
This is what traditional AI/ML is for, not GenAI.
2026-01-30 13:40:37
166
People Watcher :
Well now I want to know if Claude predicted tomorrow's travelers!
2026-01-28 21:51:40
415
LightAxeRob :
GenAI isn’t great at data analysis yet
2026-01-30 05:05:55
7
bo :
"PhD Level expertise"
2026-02-20 21:42:10
7
billerr :
I still can't fully trust that the conclusions they spit out are based on the actual data or if they're regurgitating phrasing from generally available conclusions related to my prompt. How can we verify if they actually did the work correctly without doing all of it ourselves to double-check? A conundrum
2026-01-29 21:15:17
27
TheGreatHoot :
back when covid had just gotten started, I was asked to help build a predictive model with our data scientist for tsa to try and figure out when precheck enrollment/flying would pick back up. fun times
2026-01-29 19:33:07
5
TechnicallyCreative :
step 1 should have been to come up with some ideas. step 2 would be to graph those ideas. this will give you far better results.
2026-01-29 20:35:43
6
SAM CRoSS :
This should be something to bet in polymarkets
2026-01-30 00:19:13
5
Mmmm7125 :
This is what I hate about ai. You want to offload cognitive lifting. But you see it, see it’s slightly wrong, and then have to think about it anyway so you didn’t gain anything from using it
2026-01-29 06:09:18
26
George Smith :
The reason it’s blue and orange is because it’s the default colours for a lot of python plotting packages such as matplotlib
2026-01-30 00:15:44
7
Kevin Barredo Rollie pollie :
Tuesday Wednesday Saturday. Most cheapest flight prices
2026-01-29 19:57:30
5
Vincent :
the real trend should be that business changed and people are more traveling for leisure. Which is a good take on this entire thing 🤔 interesting comparison, thanks!
2026-01-29 15:44:41
11
Saulio :
Ironically, Claud is the smartest model, but at the same is the most constrained due to less compute infrastructure compared with Open AI and Google. The response it gave you used less tokens and utilized less hardware than the others. If you had a meter data center to work on your task you would get far more intelligent results from all of them, but Claud would likely come on top. It’s good to remember that the outputs we get are not representative of the models, but the amount of compute given to them. So even if models themselves don’t improve, the consumer outputs will still improve dramatically over time due to advances in hardware and scaling data centers. This is why consumer LLMs dramatically improving over the next 10 years is an inevitability, even if scaling for training models no longer equates to better outputs.
2026-01-29 01:37:08
18
wen :
As a former corporate schmuck who was working in AI, I really think these types of exercises are incredibly important, especially to challenge our understanding of how well these AI work currently. You’ve got leadership in every single company making huge bets with people’s employment on how well these things are working.
2026-01-30 14:46:57
9
t w :
I was actually thinking this ends with massive tourism decline due to the current situation in this dangerous authoritarian developing country. How could your prediction miss this?
2026-01-29 08:03:33
7
Jacob :
It matches my travel habits almost exactly, I pretty much never fly on a Wednesday.
2026-01-29 07:12:49
42
resyeman :
Claude is so ADHD coded and I’m here for it
2026-01-31 01:59:40
7
pablo :
I believe that comparing 2019 and 2025 in day of the weak graph might show consistency in this measure (especially considering that 2019 was before Covid and a lot could have changed since then, so it is quite important to show that this is not only one year fluctuation but an existing trend)
2026-01-29 00:27:50
8
InternetJudge :
Can you do a post about " job created " under Trump? because you did an evaluation of trumps economy and i feel it was poorly representative judging by the job market.
2026-01-28 22:45:07
5
To see more videos from user @lthlnkso, please go to the Tikwm
homepage.
