And if you didn’t double check how do you know there is no AI hallucinations?
2026-03-09 15:29:29
40
mike :
He uses patterns in history to predict possible future events. He is not a prophet.
2026-03-09 21:37:31
2104
Blackbirdphys :
The raw prediction percentage isn't very informative. If someone makes many predictions of obvious things, then their score will be high. It should somehow be weighted with the prior probability of that event.
2026-03-09 16:18:43
1085
SexyTacoCat :
people have to remember, its not his conclusions that are valuable but how he gets to them.
2026-03-09 21:58:08
1103
Whoeverittakes :
This couldve been a 7 second video.
2026-03-10 05:39:25
351
gigglesmcgeee :
70% is better than most
2026-03-09 19:21:27
945
safari7 :
talked soooo much for essentially nothing
2026-03-09 20:38:33
721
that.intel.guy :
My Takeaway from this is that he is a very average analyst. We have a saying in the intelligence community and it’s, if you are more right than wrong than you are doing your job. 2/3 seems to be about right.
2026-03-09 16:43:21
96
joxer6661 :
Too much waffle. Just get to the point
2026-03-10 08:34:40
66
Agent five :
Your approach is very wrong...this is why software developers !==ai users... Ai will ignore 50% of those texts...you either need a rag system with ranking system or refine a model using those texts first....you just fed 128 hours of content and expect llm to guess with one prompt 🙄
2026-03-09 19:33:09
357
Will Legg :
I like the way your brain works
2026-03-09 15:14:09
546
Enemybird :
So Jiang is legit.
2026-03-09 15:42:33
117
Who Dat :
Considering how wild some of the topics he covers are, that's an insanely good prediction rate
2026-03-09 17:00:23
118
user5286187383667 :
his prediction is backup by facts. data . history. very educational too
2026-03-10 09:18:05
22
VK :
Keep in mind that any good predictor changes his predictors with any significant new information. That means that he constantly calibrates his prediction and he potentially changed his mind on some of predictions.
2026-03-15 00:35:24
6
MomohKouba :
That much above 50 is impressive tbf
2026-03-09 15:12:12
71
Aidan Vogel :
I've missed your content! always love your rigor on your analyses
2026-03-09 15:54:42
1
bl00dy_sabbath :
Very cool
2026-03-10 04:19:18
5
JíbaraEnTexas :
How do you know the AI isn't just lying to you ?
2026-03-09 19:48:19
72
Scootdoot :
There’s a lot of interesting ways you could continue here. For example, a prediction being right matters more if it wasn’t a popular opinion, or if it was an important prediction. If he predicted rain on Tuesday when most people said it would be sunny I wouldn’t care, but if he predicted Japan’s yen going to 0 when most thought it would strengthen that has huge implications, etc.
2026-03-09 15:17:12
42
Kizzle :
"used AI to think"
2026-03-10 15:38:03
8
Pronewbiez :
2/3s hit rate is disgustingly good if you're an investor. Munger famously said he'd be delighted with a 60% hit rate. not defending him, not am I a supporter.
2026-03-11 08:12:45
5
Jona Augustyn :
i like the way you think. ai should be used to cut through noise and compress truth, not just generate more content.
2026-03-09 20:43:21
14
Jeremy Carrasco :
Best possible use of AI in my opinion - ESPECIALLY for something so data intensive and relatively low stakes.
2026-03-09 16:44:30
5
To see more videos from user @lthlnkso, please go to the Tikwm
homepage.