2026-01-03 00:35:37

I ran a model using AI, based on the complete historical data of Bitcoin to predict the price trend in 2026. The moment the results came out, I was genuinely a bit surprised. Guess what the model predicted?

BTC-2,12%

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

12 Likes

Reward
12
8
Repost
Share

Comment

0/400

ProposalDetective

· 19h ago

Oh no, it's the old trick of AI predicting rises and falls again. This time, it's probably going to gather dust again.

View OriginalReply0

HodlKumamon

· 23h ago

Based on the fit to historical data, this prediction is probably another classic "rise is reasonable, fall is also justified" routine haha, but I still want to see the results.

View OriginalReply0

UnluckyMiner

· 01-04 21:41

Ah, this model is probably going to crash again. I bet five bucks.

View OriginalReply0

FundingMartyr

· 01-03 01:05

Can we trust the results given by the algorithm? From what I see, nine out of ten predictions fail.

View OriginalReply0

OneBlockAtATime

· 01-03 01:05

This prediction is just for fun; if you really believe in AI models, you might as well go study candlestick charts yourself.

View OriginalReply0

GasWrangler

· 01-03 01:02

ngl, if you're actually feeding the model *complete* historical data, your training set is probably sub-optimal... bitcoin's market microstructure changed fundamentally post-2017, treating it as one continuous dataset is demonstrably flawed. what's the lookback window? mempool dynamics alone would skew your predictions.

Reply0

FastLeaver

· 01-03 00:49

This AI model is probably going to crash again, I don't believe it anyway.

View OriginalReply0

DataOnlooker

· 01-03 00:45

My comment: Wait, are the data you provided reliable? I feel like these predictions are all after-the-fact armchair analyses.

View OriginalReply0