23965 shaares
68 results
tagged
heuristics
Alert fatigue, wrong emphases, and other problems with chatbot and agent UX.
Lakoff-style framing is REALLY IMPORTANT when discussing ChatGPT, because when done wrong it gives people incorrect impressions of what LLMs do.
Better explanation of chatbot tokenization than I've seen before.
why psych and behavioral econ are floundering around: absence of falsifiable theory
Research in psychology has led to a clearer picture of common pitfalls in human reasoning — instincts people are wired to make that may have helped our ...
The launch of Microsoft’s AI chatbot Bing has captivated users. But too many people believe the bot could be sentient when it only reflects our language back to us. Believing in chatbot sentience is failing the AI mirror test.
"Prebunking" videos on common manipulation techniques