Everybody’s trying to figure out how to use artificial intelligence in their company. Most of the businesses we consult with at CSG already have a patchwork of individuals experimenting with AI tools like ChatGPT; or thinking about a pilot project with some new tool or service. Most leadership teams, however…

👋 Hey friends! If you’re following the onslaught of Generative AI news like me, you’re going to want to hear about two pretty big developments that broke this past week from OpenAI and Microsoft. ChatGPT Enterprise — A Privacy-Preserving LLM for Work 🗣️ First up, OpenAI has just launched ChatGPT Enterprise, with some upgrades designed specifically for corporate…

👋 Hello friends! I’m settling into my new role as Chief Data Officer at CSG (or Cornerstone Solutions Group). We’re a consultancy that improves business operations and strategy by focusing on technology and custom software development & integration. …

I recently had a thought provoking conversation about artificial intelligence regulations with David Nicklaus, St. Louis Post-Dispatch columnist covering business & tech. The resulting article is here, which also has input from others in the St. Louis AI community like Stephen Thaler and Marc Bernstein. …

Neural networks are universal function approximaters. This means they can fit and predict very complex and arbitrary patterns, but really only work to interpolate inside the space of the data on which they’ve been trained. See below where a number of different neural networks are attempting to predict a line…

It’s interesting hearing how Accenture plans to bolster their consulting practice with $3B of investment in AI to expand their headcount, build Large Language Models (LLMs), and even make acquisitions: The large productivity gains from LLMs will disproportionately accrue to the kinds of knowledge work done in consulting (e.g…

The latest AI large language models tend to act as an intelligence equalizer. This means they provide a mathematically larger percentage gain in performance to those at the rear of the pack, bringing them closer to parity with the leaders of the pack. In a study out of MIT, 444…

Abstract How much inequality is ok? I explore this quantitatively by asking a simple question: Before financial systems existed, how were natural assets allocated? Intrinsic human traits like height, strength, & beauty were the original measures of value in which human social groups traded. Each of these attributes is distributed in…

Are large language models (LLMs) like Open AI’s GPT-4 so powerful that everyone should have fair access to them through a public utility? And if so, could their massive computational and energy requirements be served by the similarly massive compute that is simply discarded after mining Bitcoin transactions? Think of…