Intelligence Snacks podcast

Intelligence Snacks 78 - The Last Invention

0:00
1:09:25
Spola tillbaka 15 sekunder
Spola framåt 15 sekunder

Is AI humanity's last great invention? Pete and Andy question whether it's really the biggest technological breakthrough in history, comparing it with electricity, the steam engine and the internet. Is AI fundamentally different, or does every generation believe it's living through an unprecedented revolution?

They explore what intelligence means, whether AI can possess agency without desires of its own, and why humans still need to find problems worth solving. Could cheaper intelligence unleash an explosion of innovation, enable much smaller companies to tackle bigger problems, and change the value of human work?

Pete shares his latest work on Ambulando's native apps, while Andy discusses designing software through interactive canvases. They also question the predictions surrounding AGI, the future of software businesses, and whether AI might be another chapter in humanity's long history of technological change.

Key Moments:

  • [01:19] Pete's new native Mac, iPhone and Android apps.

  • [03:46] Andy's interactive canvas and the future of Figma.

  • [06:25] Why chat is a blunt instrument for AI agents.

  • [07:59] Is AI history's biggest technological breakthrough?

  • [09:47] Comparing AI with electricity and the printing press.

  • [11:12] Lessons from previous technological revolutions.

  • [12:20] Multiplying intelligence versus physical labour.

  • [14:54] Energy, hardware, regulation and AI's constraints.

  • [16:16] The human bottleneck in AI workflows.

  • [18:12] When AI agents spend too long testing.

  • [21:01] The advantage of knowing how to work with AI.

  • [22:34] Is AI just another tool?

  • [25:04] What does intelligence actually mean?

  • [27:07] Can intelligence exist without desire or agency?

  • [28:22] Why automating marketing doesn't guarantee customers.

  • [31:24] Could AI unleash an explosion of inventions?

  • [33:23] Invention versus innovation and solving real problems.

  • [35:47] Why human experience matters to problem-solving.

  • [38:52] Could embodied AI develop its own needs?

  • [39:45] Smaller companies tackling much bigger problems.

  • [41:41] Why AI slop is really about care and attention.

  • [42:52] Bullshit jobs, displacement and UBI.

  • [45:57] The human role when intelligence is on demand.

  • [49:22] Could humans leave the internet to AI agents?

  • [50:22] AGI, recursive self-improvement and magical thinking.

  • [52:13] Subordinated agency and delegating decisions to AI.

  • [56:02] Human beings, human doings and robotic work.

  • [59:37] Why original thinking becomes more valuable.

  • [01:00:38] Autonomous mining trucks and changing constraints.

  • [01:03:02] Why AI could reshape traditional SaaS businesses.

  • [01:06:57] Are we overestimating AI's historical significance?

  • [01:08:48] Why every breakthrough might feel like the last.

Projects and ideas mentioned: Ambulando, Wingman, Nostr Browser, Zap Store, Swift, Kotlin, Figma, Claude, ChatGPT, Basecamp, Salesforce, Jevons Paradox, Maslow's hierarchy of needs, AGI, recursive self-improvement, universal basic income and autonomous mining trucks.

Find out more: ambulando.io, intelligencesnacks.com

Closing thought: "If AI can't experience the problem, how is it gonna solve the problem?" — Andy

Fler avsnitt från "Intelligence Snacks"