---
date: '2024-01-07'
description: intelligence as tool for thought built atop human intelligence rather than replacing it, questioning whether next-token prediction constitutes true intelligence.
id: Intelligence amplification
modified: 2026-06-05 15:08:27 GMT-04:00
tags:
  - seed
  - ml
title: Intelligence amplification
created: '2024-01-07'
published: '2024-01-07'
pageLayout: default
slug: thoughts/Intelligence-amplification
permalink: https://aarnphm.xyz/thoughts/Intelligence-amplification.md
generator:
  quartz: v4.6.0
  hostedProvider: Cloudflare
  baseUrl: aarnphm.xyz
full: https://aarnphm.xyz/llms-full.txt
---
<article class="twitter-post" data-twitter-source="https://twitter.com/karpathy/status/1744062845426532473"><span class="twitter-post-badge" role="img" aria-label="Twitter post" title="Twitter post"><svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" viewBox="64 64 896 896" fill="currentColor" stroke="none" stroke-width="0" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true" focusable="false"><path d="M928 254.3c-30.6 13.2-63.9 22.7-98.2 26.4a170.1 170.1 0 0075-94 336.64 336.64 0 01-108.2 41.2A170.1 170.1 0 00672 174c-94.5 0-170.5 76.6-170.5 170.6 0 13.2 1.6 26.4 4.2 39.1-141.5-7.4-267.7-75-351.6-178.5a169.32 169.32 0 00-23.2 86.1c0 59.2 30.1 111.4 76 142.1a172 172 0 01-77.1-21.7v2.1c0 82.9 58.6 151.6 136.7 167.4a180.6 180.6 0 01-44.9 5.8c-11.1 0-21.6-1.1-32.2-2.6C211 652 273.9 701.1 348.8 702.7c-58.6 45.9-132 72.9-211.7 72.9-14.3 0-27.5-.5-41.2-2.1C171.5 822 261.2 850 357.8 850 671.4 850 843 590.2 843 364.7c0-7.4 0-14.8-.5-22.2 33.2-24.3 62.3-54.4 85.5-88.2z"></path></svg></span><header class="twitter-post-header"><span class="twitter-post-author">@karpathy</span><time datetime="2024-01-07">2024-01-07</time><a class="twitter-post-source" href="https://twitter.com/karpathy/status/1744062845426532473" target="_blank" rel="noopener noreferrer" aria-label="Open original post (opens in a new tab)" title="Open original post" data-skip-icons="true"><svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.5" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true" focusable="false"><path d="M7 17 17 7M7 7h10v10"></path></svg></a></header><div class="twitter-post-body">
			
				
					<p>I’m playing around with calling our tech, as it is today, IA (intelligence amplification) instead of AI. IA have the vibe of tools for thought, needing human interaction, and resemble a lot more what we actually have today. AI feels more like independent long-running agents.</p>
				
			
			
		<blockquote><cite><a href="https://x.com/simonw/status/1743785851971375468" target="_blank" rel="noopener noreferrer" data-skip-icons="true">@simonw</a></cite>
			
				
					<p>It’s OK to call it Artificial Intelligence</p>
<p>I wrote about how people really love objecting to the term "AI" to describe LLMs because they aren't actually "intelligent" - but the term AI has been used to describe this kind of research since 1955 <a href="https://t.co/wgy7lHIdWp" target="_blank" rel="noopener noreferrer">https://t.co/wgy7lHIdWp</a></p>
				
			
			
		</blockquote></div></article>

Intelligence should be thought as a tool for thought, not an independent agent

These systems should be built on top of human intelligence, not replace it.

Next-token prediction is primitive to call a system intelligent.

Can a [[thoughts/Transformers|transformers]] ever be [[thoughts/Turing-complete Transformers|Turing-complete]]?

## research area.

A lot of alpha in mechanistic analysis of the [[thoughts/representations|representations]] these models exhibit., or “virtual brain analysis”.