Episode Details

Back to Episodes
Why AI Agents Still Fail at Simple Tasks with Teng Yan

Why AI Agents Still Fail at Simple Tasks with Teng Yan

Published 3ย months, 2ย weeks ago
Description

AI promises autonomous agents that book flights and handle tasks. We're not there yet but 2026 might change everything.

In this episode of AI Supercycle, we sit down with Teng Yan of Chain of Thought to explore when AI agents will actually work autonomously and who really owns the AI powering them.

Teng breaks down the data center space race, crypto-AI infrastructure traction, and whether we're heading toward an AI privacy crisis in 2026.


We discuss:

- Why Model Improvements Are Accelerating Fast

- The Truth About Prompt Engineering

- Are We Facing an AI Privacy Crisis?

- When Will Agents Actually Work Autonomously?

- The Data Center Space Race: Bubble or Breakthrough?

- Crypto-AI Infrastructure: What's Actually Getting Traction?


Timestamps:

00:00 Intro

00:37 Near Protocol Ad

01:56 Teng's AI Background & Journey

06:15 2025: The Year AI Leveled Up

08:44 Haliday Ad, infiniFi Ad, YEET Ad

09:16 Model Improvements Across the Stack

14:22 Prompt Engineering Myths Debunked

18:45 Giving Models Context

22:10 Testing Across Multiple LLMs

24:30 AI Privacy: Crisis Brewing?

27:37 Kalshi Ad, Hibachi Ad, Trezor Ad

28:11 Near's Private AI Solution

32:40 Agent Landscape Reality Check

38:25 Why Agents Still Fail at Simple Tasks

42:15 2026: Year of the Agent?

45:30 Ethics of For-Profit AI Giants

51:20 The Data Center Bubble Question

55:45 Bit Tensor & Crypto-AI Traction

59:30 What to Watch in Blockchain AI


Website: https://therollup.co/

Spotify: https://open.spotify.com/show/1P6ZeYd...

Podcast: https://therollup.co/category/podcast

Follow us on X: https://www.x.com/therollupco

Follow Rob on X: https://www.x.com/robbie_rollup

Follow Andy on X: https://www.x.com/ayyyeandy

Join our TG group: https://t.me/+TsM1CRpWFgk1NGZh

The Rollup Disclosures: https://therollup.co/the-rollup-discl


๐——๐—œ๐—ฆ๐—–๐—Ÿ๐—”๐—œ๐— ๐—˜๐—ฅ: ๐˜๐˜ฏ๐˜ท๐˜ฆ๐˜ด๐˜ต๐˜ช๐˜ฏ๐˜จ ๐˜ช๐˜ฏ ๐˜ค๐˜ณ๐˜บ๐˜ฑ๐˜ต๐˜ฐ๐˜ค๐˜ถ๐˜ณ๐˜ณ๐˜ฆ๐˜ฏ๐˜ค๐˜บ ๐˜ข๐˜ฏ๐˜ฅ ๐˜‹๐˜ฆ๐˜๐˜ช ๐˜ฑ๐˜ญ๐˜ข๐˜ต๐˜ง๐˜ฐ๐˜ณ๐˜ฎ๐˜ด ๐˜ค๐˜ฐ๐˜ฎ๐˜ฆ๐˜ด ๐˜ธ๐˜ช๐˜ต๐˜ฉ ๐˜ช๐˜ฏ๐˜ฉ๐˜ฆ๐˜ณ๐˜ฆ๐˜ฏ๐˜ต ๐˜ณ๐˜ช๐˜ด๐˜ฌ๐˜ด ๐˜ช๐˜ฏ๐˜ค๐˜ญ๐˜ถ๐˜ฅ๐˜ช๐˜ฏ๐˜จ ๐˜ต๐˜ฆ๐˜ค๐˜ฉ๐˜ฏ๐˜ช๐˜ค๐˜ข๐˜ญ ๐˜ณ๐˜ช๐˜ด๐˜ฌ, ๐˜ฉ๐˜ถ๐˜ฎ๐˜ข๐˜ฏ ๐˜ฆ๐˜ณ๐˜ณ๐˜ฐ๐˜ณ, ๐˜ฑ๐˜ญ๐˜ข๐˜ต๐˜ง๐˜ฐ๐˜ณ๐˜ฎ ๐˜ง๐˜ข๐˜ช๐˜ญ๐˜ถ๐˜ณ๐˜ฆ ๐˜ข๐˜ฏ๐˜ฅ ๐˜ฎ๐˜ฐ๐˜ณ๐˜ฆ. ๐˜ˆ๐˜ต ๐˜ค๐˜ฆ๐˜ณ๐˜ต๐˜ข๐˜ช๐˜ฏ ๐˜ฑ๐˜ฐ๐˜ช๐˜ฏ๐˜ต๐˜ด ๐˜ต๐˜ฉ๐˜ณ๐˜ฐ๐˜ถ๐˜จ๐˜ฉ๐˜ฐ๐˜ถ๐˜ต ๐˜ต๐˜ฉ๐˜ช๐˜ด ๐˜ค๐˜ฉ๐˜ข๐˜ฏ๐˜ฏ๐˜ฆ๐˜ญ, ๐˜ธ๐˜ฆ ๐˜ฎ๐˜ข๐˜บ ๐˜ฆ๐˜ข๐˜ณ๐˜ฏ ๐˜ข ๐˜ค๐˜ฐ๐˜ฎ๐˜ฎ๐˜ช๐˜ด๐˜ด๐˜ช๐˜ฐ๐˜ฏ ๐˜ฐ๐˜ณ ๐˜ง๐˜ฆ๐˜ฆ ๐˜ข๐˜ด ๐˜ข ๐˜ด๐˜ฑ๐˜ฐ๐˜ฏ๐˜ด๐˜ฐ๐˜ณ๐˜ด๐˜ฉ๐˜ช๐˜ฑ, ๐˜ช๐˜ง ๐˜ต๐˜ฉ๐˜ช๐˜ด ๐˜ช๐˜ด ๐˜ต๐˜ฉ๐˜ฆ ๐˜ค๐˜ข๐˜ด๐˜ฆ ๐˜ธ๐˜ฆ ๐˜ธ๐˜ช๐˜ญ๐˜ญ ๐˜ข๐˜ญ๐˜ธ๐˜ข๐˜บ๐˜ด ๐˜ฎ๐˜ข๐˜ฌ๐˜ฆ ๐˜ด๐˜ถ๐˜ณ๐˜ฆ ๐˜ช๐˜ต ๐˜ช๐˜ด ๐˜ค๐˜ญ๐˜ฆ๐˜ข๐˜ณ. ๐˜ž๐˜ฆ ๐˜ข๐˜ณ๐˜ฆ ๐˜ด๐˜ต๐˜ณ๐˜ช๐˜ค๐˜ต๐˜ญ๐˜บ ๐˜ข๐˜ฏ ๐˜ฆ๐˜ฅ๐˜ถ๐˜ค๐˜ข๐˜ต๐˜ช๐˜ฐ๐˜ฏ๐˜ข๐˜ญ ๐˜ค๐˜ฐ๐˜ฏ๐˜ต๐˜ฆ๐˜ฏ๐˜ต ๐˜ฑ๐˜ญ๐˜ข๐˜ต๐˜ง๐˜ฐ๐˜ณ๐˜ฎ, ๐˜ฏ๐˜ฐ๐˜ต๐˜ฉ๐˜ช๐˜ฏ๐˜จ ๐˜ธ๐˜ฆ ๐˜ฐ๐˜ง๐˜ง๐˜ฆ๐˜ณ ๐˜ช๐˜ด ๐˜ง๐˜ช๐˜ฏ๐˜ข๐˜ฏ๐˜ค๐˜ช๐˜ข๐˜ญ ๐˜ข๐˜ฅ๐˜ท๐˜ช๐˜ค๐˜ฆ. ๐˜ž๐˜ฆ ๐˜ข๐˜ณ๐˜ฆ ๐˜ฏ๐˜ฐ๐˜ต ๐˜ฑ๐˜ณ๐˜ฐ๐˜ง๐˜ฆ๐˜ด๐˜ด๐˜ช๐˜ฐ๐˜ฏ๐˜ข๐˜ญ๐˜ด ๐˜ฐ๐˜ณ ๐˜ญ๐˜ช๐˜ค๐˜ฆ๐˜ฏ๐˜ด๐˜ฆ๐˜ฅ ๐˜ข๐˜ฅ๐˜ท๐˜ช๐˜ด๐˜ฐ๐˜ณ๐˜ด.

Listen Now

Love PodBriefly?

If you like Podbriefly.com, please consider donating to support the ongoing development.

Support Us