Unfiltered AI Talk: "AI President? Our AI Isn't Advanced Enough" | Episode [36]

3 Nov 2025 · 2 min
2:05

Speaker (creator) explains they are building an open-source personal AI that exposes what the system knows about a user and allows the user to correct or remove facts (example: the AI incorrectly reminding the speaker about a deceased dog). They argue this transparency and user control is necessary because commercial systems are opaque and collect/monetize user data; they view deep integration of AI into daily life as inevitable and emphasize safety, guardrails, ethics, and user-facing tools. The speaker asks viewers to follow along for an upcoming release in the next few weeks and solicits feedback.

Questions this video answers

Yes — the speaker says "our stuff is all open source."

"In the next couple of weeks."

More from Robert

1:00:59
youtube.com

What OpenAI Dev Day 2026 Means For AI Builders

2d ago
4:08
youtube.com

Why Max Reasoning Is Usually a Waste

2d ago
2:57
youtube.com

Why You Don't Need Frontier AI for Everything

2d ago
0:44
youtube.com

Max reasoning can make AI models worse

2d ago
3:01
youtube.com

How Cursor Rebuilt SQLite for $1,339 Instead of $20,000

3d ago
0:57
youtube.com

You don't need frontier AI for everything

3d ago
1:22
youtube.com

Cursor rebuilt SQLite for $1,339 instead of $20,000

4d ago
21:00
youtube.com

Cursor Just Showed How Much You're Wasting On AI

4d ago
3:22
youtube.com

How I Made a Product Video With One Prompt

4d ago
5:55
youtube.com

How To Cut the Cost of LLM Judges

5d ago
1:15
youtube.com

Anthropic's Opus 5 flop was a people problem

5d ago
See all content