AI has a lying problem —
https://astralcodexten.substack.com/p/elk-and-the-problem-of-truthful-ai

Why Does AI Lie, and What Can We Do About It?
AI Alignment.
AI has a lying problem.

“Make up 20 words, and provide a definition along with each made-up word.”

Confirmed — You can poll GPT just like humans, by asking it the same question many times, and GPT will respond shockingly similarly to how humans would respond

https://arxiv.org/abs/2209.06899

“When provided with real survey data as inputs, GPT-3 reliably answers closed-ended survey questions in a way that closely mirrors answers given by human respondents. The statistical similarities extend to a whole set of inter-correlations between measures of personal behaviors, demographic characteristics, and complex attitudes. We again see this as strong evidence for algorithmic fidelity”

Social science about to change.
“among the most significant things I’ve ever covered”

trlX allows you to fine-tune language models using reinforcement learning via either a provided reward function or reward-labeled dataset.

https://github.com/CarperAI/trlx

trlX AI alignment

“write a smart sounding tweet about macro economics and impact of the dot plot peaking at 5.1% on bitcoin”

No-code way to launch AI-powered apps

Demo: https://www.youtube.com/watch?v=0cms4b7UqJg

Waitlist: https://www.getaspenai.com

Evil AIs of Fiction