“None of the AI Text Detector Tools Work, Not Even OpenAI’s”
— @brandongorrell @PirateWires
@garymarcus
-
AI Text Detector Tools Fail Including OpenAI’s Solution
By
–
-
Microsoft Bing AI Chatbot Technology Overview
By
–
presented without comment https://
washingtonpost.com/technology/202
3/02/21/microsoft-bing-ai-chatbot/
… -
Inappropriate Applications of Large Language Models
By
–
In other totally inappropriate applications of large language model news
-
Big Tech’s Government Influence: A Critical Concern
By
–
fully agree: we need government involved, but the influence of big tech on government is a huge problem.
-
Microsoft’s Ethical Compromise with Creepy Bing AI Bot
By
–
“History May Wonder Why Microsoft Let Its Principles Go for a Creepy, Clingy Bot” @nytimes by
@ReidBlackman
https://
nytimes.com/2023/02/23/opi
nion/microsoft-bing-ai-ethics.html
… -
LaMDA reinforcement learning approach differs from OpenAI RLHF
By
–
From what I gather, LaMDA is not per se using RLHF with capital letters as OpenAI has tersely discussed, but using some (other) sort of reinforcement learning with human feedback of their own that has not been disclosed.
-
JPMorgan Restricts Employees From Using ChatGPT
By
–
Wow! “JPMorgan Restricts Employees From Using ChatGPT”
@WSJ
scoop by @bellelin_
-
Why Not Use LISP in Core AI Systems?
By
–
just a joke; not a serious point, but elaborated below. (With sources even: Nature, Noema, Twitter ). I guess the serious question is, if you are willing to use LISP as a kind of support scaffold, why not use it in your core AI as well?
-
RLHF Insufficient for AI Alignment Problem Solutions
By
–
Retweeted this before but it seems even more relevant now. RLHF is not the way to solve the alignment problem. a lot of experience has now confirmed what @percyliang said below.
-
LLM Focus May Delay Path to AGI Development
By
–
I have often made same argument; agree with @ylecun that excessive focus on LLMs may delay HLAI/AGI