- Published on January 23, 2023
- In AI Features
Human Feedback Frenzy: How it Turns AI into Narcissistic, Control-Freak Machines
“The path I'm very excited for is using models like ChatGPT to assist humans at evaluating other AI systems,” said OpenAI’s Jan Leike
- By Ayush Jain
OpenAI's GPT-3.5 architecture, which runs ChatGPT, is equipped with reinforcement learning from the human feedback model (RLHF), a reward-based mechanism based on human feedback to improve its responses. Essentially, one can suppose that the chatbot is trained in real time by human inputs.
However, the RLHF system has also had its own set of consequences. Sarah Rasmussen, a Cambridge University mathematician, gave the following example to show that the model favours being rewarded for achieving a desired outcome rather than having a definite idea of what is right.
https://twitter.com/SarahDRasmussen/status/1609972620761473027
This is not just a one-off case. To test it further, we asked ChatGPT for the name of the current CEO of Twitter. In the first instance, it did
📣 Want to advertise in AIM? Book here
Ayush Jain
Ayush is interested in knowing how technology shapes and defines our culture, and our understanding of the world. He believes in exploring reality at the intersections of technology and art, science, and politics.
Related Posts
How This Coimbatore SaaS Firm Cracked Hidden Enterprise Problem Costing Millions
Founded in 2015, now based in Portland, Responsive supports more than 20% of Fortune 100 companies
Enterprises Beware: Agent-Washing Clouds the Future of AI
Vendors mislabel copilots as agents, raising regulatory and operational risks for firms chasing the promise of agentic AI.
How Neysa Stands Out in the IndiaAI GPU Race
Unlike other providers focused on GPU allocation, Neysa claims to deliver an end-to-end AI cloud platform.
Two Indian Engineers on a Mission to Automate Home Cooking for the World
In a live demonstration for AIM, Posha prepared paneer tikka masala in approximately 25 minutes
BharatGen and the Pursuit of Sovereign, Scalable AI for India
“Knowledge-driven components are important because we don’t want everything to be just algorithmic innovation.”
How Pradhi AI Embeds Emotional Intelligence in Voice AI
As businesses recognise the potential of voice-driven tech, Pradhi AI is laying the foundation for an empathetic, responsive AI ecosystem.
Mangaluru Looks to Build Its Own Tech Identity, Not Replicate Bangalore
“The coastal city could showcase tangible results by applying deep tech to areas it already dominates”
Google’s Gemini Nano Banana and the Cost of Convenience
The company’s new AI image and photo editor deepens concerns over data use and consent gaps, experts warn.
Download the easiest way to
stay informed
stay informed
