paul-larsen-data-ai.com

I’ve given a few examples in the past of how a prompt (natural language instruction) to a large language model is just a suggestion and not a rule.

One of the many reasons I loved the IT security event, Ljubljana BSides, was the chance to hear (and give) lightning talks at the end.

(Slovenska verzija) Umetna inteligenca lahko seveda ponudi korist podjetjem, vendar prinaša tudi tveganja.

A big thank you to Thomas Albermann, is a Senior Associate specializing in data and AI regulation at Bird & Bird, one of the top tech (and more) law firms out there.

What do you do with a consumer product that poses danger to its users?

A big thank you to Ksenija Jaklin and Matej Krumberger of the Slovenian Banking Association, Združene Bank Slovenija, on AI Risk Management.

I was using Anthropic’s Claude chat this morning to generate some utility code to evaluate how well different algorithms work for the task of extracting tabular data from semi-structured documents, like invoices, utility bills or financial reports.

From productivity dreams to cyber risk nightmares The impacts of GenAI on business IT are real.

Two days in a row of gentle, I hope constructive, rants about the power and perils of GenAI.

I discovered Edwin McCain’s music on a live National Public Radio show once driving home in high school, and he’s still one of my favorites.