Long article on the more recent history of AI - and why it seemed to explode after having been a "backwater" for so long. Why the deep learning boom caught almost everyone by surprise (bit.ly/4hHNgtS). Interesting to know the history...

This is a bit odd, Amazon’s CEO defends return-to-office policy (bit.ly/3Z7aaUx) and says they are defending the culture. Most of the "culture" I have heard from AWS staff is that it is a "sweat shop". I guess you can't overwork people remotely....

MIT research finds that Despite its impressive output, generative AI doesn’t have a coherent understanding of the world (bit.ly/3NZpBrn). They don't really understand the world, but they do have a great way of articulating things. They have their uses.

I know, the author sounds like an IBM fanboy, but a lot of the content in here is thought provoking. IBM Granite 3.0 Models Point The Way To Enterprise AI At Scale (bit.ly/4f5eAk9). Granite is more cost efficient and makes business models for some approaches a reality.

The AI hype train is at full roar... but an old friend of mine just finished up a CIO Insourcing project. (ibm.co/3UCP5yk). Rosalind and her folks just moved IT workload into consolidated data centers and saved IBM millions of dollars of real money. No hype - results.

Data Centers for AI are so hungry that they require their own nuclear power plants. There's a punchline in here somewhere... Meta nuke powered AI data center (bit.ly/3UDrPQJ) gets stopped by discovery of rare bee species. Interesting trade offs being made in this space.

I've never been a big user of "mind maps", I find that they often take my focus and put it on connecting the boxes rather than connecting the concepts. However, I may take Shay's advice here and try this new approach to building mind maps (bit.ly/40wwzvu).

More than you ever wanted to know about timezones and calculating the time where you live. Learn about how Australia/Lord_Howe is the weirdest timezone (bit.ly/3AAmkLQ), and be glad that someone else deals with this and figures it out for us.

When I began in Software Development, it was a solitary discipline, that few people understood. Now I find that SW development is more of a team sport, as described On Good Software Engineers (bit.ly/4ftkLhF), where teamwork is essential.

When I read Get Me Out Of Data Hell (bit.ly/3UujyyM), I just took some deep breaths and remembered some of the horrors that I have witnessed. Tools like Cloud Pak for Data can help here, but tools alone don't fix the problem. You need a commitment to managing your data

I find myself slowly coming back to the fundamentals of SW development, when I work with AI. The Part of PostgreSQL We Hate the Most (bit.ly/4fnvISx). MVCC has been around for a long time... I'm thinking of all of the ClearCase installations that I deployed...

Saw this guide to Grokking Algoritms (bit.ly/3YtVk8L) and immediately went back to my old high school days when I learned about sort complexity, quicksort, binary sorts, hashes, and all of the common tools in the developer toolkit. This is a cool little guide.

Introducing SimpleQA (bit.ly/3UuMC9c), a new AI benchmark for frontier models. The results are not pretty. Less than 50% accuracy for any model, most are far worse than that. Maybe we should ask these models what a "stochastic parrot" is...

I don't agree with everything in this post, I’m Now Terrified of AI, And You Should Be Too (bit.ly/40jxuzg), but I do like the points about the costs of Generative AI when associated with its reliability, and AI's inability to support mission critical computing jobs.

Some interesting observations from an old friend on the new watsonx Granite models (bit.ly/3AhhXW4). Always interesting to see what Ramesh is doing and thinking about.

Saw this post from Bill Higgins (bit.ly/48zTub9), and read the article that this pointed to (bit.ly/3Aahx3c). It's a pretty cool little example of an open source co-pilot that ANYONE can build and use.

This is a bit of a "hot take" on social media and the relative safety of children online. Some of the observations are a bit jarring, but they do drive home the authors points. Thanks to Ryan Anderson for pointing me towards this article on Age Gating (bit.ly/4hj5mSH)

An interesting viewpoint and history on the journey to becoming a technical leader. The Path from Senior Engineer to CTO (bit.ly/4f1fNZi) has some good tips for people who think that they are prepared, both in terms of emotion and personality, and in technical experience

A lot of our software infrastructure is based on open source technologies. How do we fund open source? (bit.ly/3BMSBji). How do we make sure that open source is secure? It's not just about paying people, it's about removing temptation to act in ways you wouldn't want.

Microsoft Azure CTO: US data centers will soon hit size limits (bit.ly/4dVoAeA). I think that people would be shocked if they knew the amount of money and energy that is routinely consumed for AI and cryptocurrency. It has a definite impact - and a definite cost