Quick Reads (Under 5 Minutes)

The high quality data that is used to train models is becoming monetized. That is inevitable. Society will have to arrive at some policy for uses that genuinely lie in the public interest. But commercial model builders should pay and so will us commercial consumers.

‘The researchers estimate that in the three data sets — called C4, RefinedWeb and Dolma — 5 percent of all data, and 25 percent of data from the highest-quality sources, has been restricted. Those restrictions are set up through the Robots Exclusion Protocol’

If gpt 4o-mini is really less energy consumptive and has some of the characteristics of a more consistent, straightforward automation friendly model like claude sonnet, then this release might be a big win for those of us ”lucky” enough to spend part of our time in the Azure OpenAI ecosystem.

Fingers crossed. Skepticism unquelled.

My main hope for the future of AI is that we use it to make our world a little more fair for people.

AI is capable of contributing to our worst nightmares but only if we make it so. It can also help us reduce scarcity and democratize access.

History shows we are capable of moderating ourselves. And in that space allowing empathy and integrity to inform our actions. So. I hope. As a choice.

My article at the top of the infoq roundup. My ego wants me to believe it’s not just really good micro targeting but that other people are finding it useful. https://lnkd.in/eqF32cjN

I apply pair programming experience to how I engage code assist. I carry on a reasoning conversation. I ask questions of clarification. I make design decisions. I alternate between tests and code. And, I write code.

I still value pairing with coworkers. Especially on interesting problems.

I also appreciate being on a channel with my coworkers while we work on our own things.

I take great pleasure and learn from all of it. Most importantly for others, I move quickly and cross gaps in domain knowledge and language familiarity while exercising craft.

In short, I still reside in a world where things of practical beauty are made by people who take pride and joy in their work.

Taking a pause to reflect on the joys of collaborating in small numbers. I’ve been working closely with one coworker who I respect and whose capabilities complement mine.

The speed and clarity with which we can move. The level of tacit communication and mutual support. The learning and freedom I feel in my work is inspiring and exceptionally fun.

I’ll enjoy it while it lasts because like all experiences it is momentary stop in a career that has become nothing but transition and change both in terms of the opportunities and my own wants and aspirations.

Fortunate to work with a team that wants to spend more time on retrospection. I say that knowing they spend the time well:

• time for reflection and learning,
• probing analysis on single concerns
• select set of things to try
• that are followed up on in subsequent sessions.

A career isn’t a track. It’s a stream. It runs then jumps its bank and cuts a new course. A year ago, I wouldn’t have guessed where I find myself now. I flow with it. Try to influence its direction and take in the moments as I can.

Reposting. If you are commited to leadership development and have some resources to send experienced and new leaders to a learning event, particularly if you are in the NY, NJ, CT area. I encourage you to consider attending July 9-12 in Montclair.

https://lnkd.in/eizrkRRQ

https://lnkd.in/eHakUdrc

Using AI agents to help summarize, categorize, and analyze customer pain points. From my infoQ article on Using Generative AI in Software Project Management . https://lnkd.in/eqF32cjN

Ironic..

I built a dependency tracer in C# which traces method invocation through a codebase. It produces a manifest of the relevant code as retrieval for RAG and also produces plant UML descriptors to create class and sequence diagrams

Running it against legacy code, I get a clean path from input source to database. Running it against my own tracer code, I get nothing useful.

Static analysis is not good at tracing runtime interactions like dependency injection and implementations behind interfaces. ie well architected code! After banging my head against it for a few hours, I realized it kind of works. If it finds nothing to rewrite, it found nothing worth rewriting…