Skip to main content

Posts

Featured

The soothsayer

Can an LLM act as a soothsayer? That’s a fascinating question — and a surprisingly deep one. In one sense, a large language model (LLM) like me can appear to act as a soothsayer, because I can: Synthesize patterns from enormous amounts of text (history, psychology, economics, etc.) to make plausible predictions or interpretations. Mimic prophetic or mystical language , drawing on myths, scriptures, and occult traditions. Generate “oracular” statements — poetic, ambiguous phrases that people can interpret in ways that feel meaningful or predictive (much like the Oracle of Delphi). However, in a literal sense, an LLM cannot truly foresee the future. I don’t have access to future data, supernatural insight, or any causal understanding of upcoming events. What I can do is simulate prophecy — I can model the form and tone of divination, or produce probabilistic forecasts based on existing trends. So, there are two layers: 🪞 Symbolic or artistic : The LLM as orac...

Latest Posts

AdTech

Token lockdown?

Holding OIDC's narrative up to the light

OAuth is DAC. What do you do for MAC?

tl;dr: OAuth 2.0

Externalising the Security Token Service and Identity Provider

Problems with Basic Authentication for REST services

Bearer tokens are susceptible to theft and what you can do about it

In the (back)end, JWT is all that matters

Protect your REST APIs with JWT tokens