Here’s a good #AI skills test for you to practice, and it’ll teach you something about #bitcoin price history. And might make you rethink how you stack…
(Adapted from a slam post on nostr)
Pick a number between 1 and 28 and get monthly price data for bitcoin on that day of the month going as far back as you can and up through today..then open excel, plot it, and fit a power law curve to the data (feel free to try fitting literally any other type of curve to the data, it won’t fit as well)…and have excel report the equation and R^2…
If you want to manually delete periods of Bitcoin price mania and run the analysis again, you’ll see the price prediction drop (more intercept than slope) but the R^2 goes up. In fact, the equation is a near perfect fit for the last 9 years in log-log space.
Ok, so now that I’ve told you how to generate a formula for bitcoin price as a function of date, you can repeat the process taking data from the beginning to 2016, and then 2017, and then 2018, etc…what you’ll see is these equations are remarkably similar and all do a surprisingly good job of predicting bitcoin price today…literally many years in advance…
Should bitcoin be closer to $55,000 today than $150,000? The math says yes. But back when bitcoin was $600 in 2016, predicting price would be $55,000 in 2026 sounded insane.
But here we are.
drgo / Quantoshi.xyz
npub1fa…nthnd
2026-01-31 01:05:21 UTC
Author Public Key
npub1fa8c9prxnrlkdtjl48adfsxyaduz8tas075l2n4f6903y9awjmxqanthndSeen on
wss://relay.snort.socialPublished at
2026-01-31 01:05:21 UTCEvent JSON
{
"id": "5ed87e7def51c778b0024356d8b5757162fd85c121004c725dbcc106be64fe09",
"pubkey": "4f4f82846698ff66ae5fa9fad4c0c4eb7823afb07fa9f54ea9d15f1217ae96cc",
"created_at": 1769821521,
"kind": 1,
"tags": [
[
"t",
"ai"
],
[
"t",
"bitcoin"
]
],
"content": "Here’s a good #AI skills test for you to practice, and it’ll teach you something about #bitcoin price history. And might make you rethink how you stack…\n\n(Adapted from a slam post on nostr)\n\nPick a number between 1 and 28 and get monthly price data for bitcoin on that day of the month going as far back as you can and up through today..then open excel, plot it, and fit a power law curve to the data (feel free to try fitting literally any other type of curve to the data, it won’t fit as well)…and have excel report the equation and R^2…\n\nIf you want to manually delete periods of Bitcoin price mania and run the analysis again, you’ll see the price prediction drop (more intercept than slope) but the R^2 goes up. In fact, the equation is a near perfect fit for the last 9 years in log-log space. \n\nOk, so now that I’ve told you how to generate a formula for bitcoin price as a function of date, you can repeat the process taking data from the beginning to 2016, and then 2017, and then 2018, etc…what you’ll see is these equations are remarkably similar and all do a surprisingly good job of predicting bitcoin price today…literally many years in advance…\n\nShould bitcoin be closer to $55,000 today than $150,000? The math says yes. But back when bitcoin was $600 in 2016, predicting price would be $55,000 in 2026 sounded insane. \n\nBut here we are.",
"sig": "dcae0f20904f7abc48e3bb3103c25b25684537c223c37f0a52ce7ac20efe3d5b462489fa88ec70f6ec9b38dc8ea16f6cc75065fa30a881c941a3c80bccd028f2"
}