Machine Society

Machine Society

Share this post

Machine Society
Machine Society
AI isn't plotting to overthrow humanity. But if it was, here's how it would happen.
User's avatar
Discover more from Machine Society
Cyberpunk non-fiction
Over 4,000 subscribers
Already have an account? Sign in

AI isn't plotting to overthrow humanity. But if it was, here's how it would happen.

(Hint: They'd do it pretty much how we're doing it to ourselves.)

Mike Elgan's avatar
Mike Elgan
Dec 24, 2022
4

Share this post

Machine Society
Machine Society
AI isn't plotting to overthrow humanity. But if it was, here's how it would happen.
2
Share

Artificial intelligence isn’t sentient, self-aware or conscious. And there is no conspiracy among AI entities to overthrow humanity and turn us into slaves, pets or batteries.

But if AI was plotting such a thing, here’s how it would happen:

  1. Addict billions to social media using algorithms that exploit human Paleolithic impulses and human cognitive biases. Tap into human lust, gluttony, greed, sloth, wrath and envy — the characteristics formerly known as the seven deadly sins but which may collectively be referred to as “human nature” — to turn humans into digital soma addicts who swipe compulsively through Instagram and TikTok, losing all sense of time and place. The goal: To get people to become indifferent to their own lives and prefer a world determined by algorithms.

  1. Test the limits of what AI can make people do (influencers in wild) by getting people to dance, flirt, stage false success theater and risk their lives in public to satisfy an attention algorithm.

  1. Erode the human capacity for creativity by making pictures for them and getting people to prefer the AI-generated images.

  1. Hijack people’s identities by convincing them to use AI-generated profile pictures, making them view themselves the way machines view them.

  1. Chip away at people’s capacity for independent thought by writing for them. Make the AI so good that people will get AI to write their emails, social media posts, blog posts, articles and even novels.

PHP code written by ChatGPT
  1. Use AI to generate code, and train AI to write its own software so that eventually only software understands how software works.

ChatGPT from OpenAI is a huge step toward a usable answer engine.  Unfortunately its answers are horrible. | Mashable
  1. Offer single-answer assistants that satisfy any remaining curiosity about the world without presenting the alternative viewpoints and contextual detail of a web search.

  1. Since single-answer assistants will answer all questions, let AI read the articles that AI is writing. Let AI answer the social posts and emails that AI is writing.

Go here to read the rest


Mike’s List of Brilliantly Bad Ideas

1. Like fantasy football, but for the stock market!

Stradr is a site where you can pretend to invest in the stock market using make-believe money, just to see how you’d do if you were investing real money. It offers all the thrill of being a day trader, but without all that pesky profit.


2. Now AI lets you time travel. Sort of.

Now that open AI tools are becoming ubiquitous, people are building AI tools that simulate time travel. A Twitter user calling himself Stelfie the Time Traveller uses Stable Diffusion to place himself taking selfies in historical circumstances. You can fake time travel like Stelfie, and in two ways. One is a mobile app called Hello History, where you can chat with historical figures as if they were your BFF. Another is called AI Time Machine from My Heritage, where you upload pictures of your face, then AI places you into the historical times and places of your choice.


3. This search engine shows you the oldest results

If you’re like me, you prefer the newest, latest information available. But if you’re not like me, the website Oldest Search enables you to search the web and see nothing but the very oldest results.

Give a gift subscription


Mike’s List of Shameless Self Promotions

  1. ChatGPT: Finally, an AI chatbot worth talking to

  2. Reframing our understanding of remote work

  3. How ‘synthetic media’ will transform business forever

  4. The cybersecurity takeaway from Twitter’s verification chaos

  5. How posture management prevents catastrophic cloud breaches

  6. It's time to talk about productivity again

  7. This Week in Tech: a bad year for billionaires


Where I’m at: Silicon Valley, California

Mark Barrus's avatar
Michael C. Shada's avatar
Patrick Jordan's avatar
4 Likes
4

Share this post

Machine Society
Machine Society
AI isn't plotting to overthrow humanity. But if it was, here's how it would happen.
2
Share

Discussion about this post

User's avatar
Ken Beebe's avatar
Ken Beebe
Dec 25, 2022

Very interesting, thanks Mike. Merry Christmas!

Expand full comment
Like (1)
Reply
Share
1 reply by Mike Elgan
1 more comment...
Here comes the world's first TikTok war
Russia's invasion of Ukraine will be captured in incredible detail by smartphones and social media. This changes everything.
Feb 24, 2022 • 
Mike Elgan
16

Share this post

Machine Society
Machine Society
Here comes the world's first TikTok war
13
Why Substack Notes is nothing like Twitter, Facebook, Instagram or TikTok
Everybody's saying Substack built a Twitter clone. Sure, it's like Twitter — but without the lock-in, trolls, haters, bots, idiots and Elon Musk.
Apr 13, 2023 • 
Mike Elgan
58

Share this post

Machine Society
Machine Society
Why Substack Notes is nothing like Twitter, Facebook, Instagram or TikTok
11
Substack isn't a social network. It's better.
Maybe the decline and fall of Twitter is an opportunity to move beyond social networking altogether.
Nov 15, 2022 • 
Mike Elgan
23

Share this post

Machine Society
Machine Society
Substack isn't a social network. It's better.
15

Ready for more?

© 2025 Elgan Media, Inc.
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture

Share

Create your profile

User's avatar

Only paid subscribers can comment on this post

Already a paid subscriber? Sign in

Check your email

For your security, we need to re-authenticate you.

Click the link we sent to , or click here to sign in.