3b‐1 My take on agents - terrytaylorbonn/auxdrone GitHub Wiki

26.0317 Lab notes (Gdrive) Git

This page is for now mainly notes about the AI agent frameworks (like Harness, etc). The latest dev directions. Will start doing demos after return to USA 26.0326.

  • 1 The evolution of my AI focus
  • 2 Input junk into a transformer/CNN and you get junk output. Intelligence?
  • 3 End users need a realistic (honest) assessment of AI
  • 4 AI will never have intelligence, so how to move forward? Limit AI inputs
  • 5 My demo roadmap
  • 6 My Substack posts
  • 7 Youtube videos

1 The evolution of my AI focus

  • Phase 1 drones (with CNN) (2.5 years ago).
  • Phase 2 LLMs.
  • Phase 3 robotics AI (early 2026).
  • (now) Phase 3b AI agents (I was working a year ago primarily on Langchain; AI agents have become much more interesting lately).
image



2 Input junk into a transformer/CNN and you get junk output. Intelligence?

  • Input junk and AI outputs the most probable match.
  • Input something thats not junk, but for some reason the UFA does not approximate correctly, then the result is junk.
  • AI does not warn you when it outputs junk.
image



3 End users need a realistic (honest) assessment of AI

I have been focusing a lot recently on my long-standing claim that GPU-based AI has no intelligence and never will. The reasons for this

  • 1 Honesty pays (especially if you are an AI end-user).
  • 2 AI gurus know that AI has no intelligence.
  • 3 To truly exploit AI, you need to have a realistic assessment of its capabilities.

Slowly the limitations of AI (it is not intelligent and never will be) are being acknowledged. Part of the reason is the rapidly growing customer base is accumulating a lot more hands-on experience.



4 AI will never have intelligence, so how to move forward? Limit AI inputs

  • Create a “sandbox” for the LLM that will limit the input to the LLM.
  • This will be analogous to binary apps running on an OS “sandbox” (preventing the app from crashing the OS).
image



5 My demo roadmap

This is still under development. I will start on this on 26.0326. Wiki page: https://github.com/terrytaylorbonn/auxdrone/wiki/3b%E2%80%902-Demo-roadmap .  



6 My Substack posts



7 Youtube videos

Recent videos from my favorite AI blogger “Best Partner” that sum things up perfectly.

  • 7.1 A VC analysis of the future agent market (it belongs to those who adapt quickly)
  • 7.2 The future of AI belongs to framework tools that maximize AI model reliability (such as Harness)
  • 7.3 An AI guru talks about the “bitter lesson” (bitter for AI gurus) of the unrealized promises of AI
  • 7.4 An AI guru who has far more modest expectations (promises) for AI  

7.1 A VC analysis of the future agent market (it belongs to those who adapt quickly)

https://www.youtube.com/watch?v=mNhHjkf_ahM 【人工智能】Cursor要凉了么 | 顶级风投杰瑞·默多克警告过时 | 自主Agent | SaaS海啸主浪潮未至 | 编排层革命 | 软件买家剧变 | 基础设施重构 | AI原生公司 | 科技趋势

image



7.2 The future of AI belongs to framework tools that maximize AI model reliability (such as Harness)

【人工智能】Agent Harness Engineering | Agent驾驭/管控工程 | 长时任务的缺陷 | 计算机的操作系统 | 通用型和垂直型 | 苦涩的教训 | 工程实践 https://www.youtube.com/watch?v=qua6FfJmydo

image



7.3 An AI guru talks about the “bitter lesson” (bitter for AI gurus) of the unrealized promises of AI

The following is a quote from Richard Sutton (AI guru) in the recent YT video “Can humans make AI any better?” (from Welch Labs https://youtu.be/2hcsmtkSzIw?t=547 ). This article states that it is a bitter lesson to discover that the “contents” of minds are “irredeemably” complex (I think he means “intelligent”).

The answer: Use the same tech that failed to created intelligence to instead search and discover the world by itself (not sure if this means visually, sound, touch, etc). This will somehow solve the problem of a lack of intelligence (??; I don’t quite understand the logic).

image



7.4 An AI guru who has far more modest expectations (promises) for AI

Perhaps not coincidentally he does not make money or get investment directly from AI (his main projects now are educational).

【人工智能】AGI还早着呢 | 吴恩达 | Agentic AI | 规模化Scaling Laws瓶颈 | 持续学习难题 | 灾难性遗忘 | 开源模型优势 | 赋能人类 | 职业重构 | 行业泡沫

https://www.youtube.com/watch?v=D5rlQiSvbek

image



⚠️ **GitHub.com Fallback** ⚠️