I have known and been attending events put on by Eric and Kim Norlin for over 15 years – Defrag, Glue, and now SW2con – you can see the agenda here. Each event series (my words) focuses on something big happening in the world of technology – mobile, machines, data, social, cloud, APIs and open architectures….and now artificial intelligence from the enterprise software development point of view vs. hype and bluster. These events are intentionally small and held at the Omni hotel in Broomfield, Colorado so brings folks in from all over including both coasts as well as local brainpower.

Over the years as these conferences have become more technical (a point of pride if you ask Eric). I often feel like I am wasting a seat as I am more “of technology” than technologist but, in my mind, it is a cheat for my job now as an investor to see what is real, what is coming as well as meet some very, very smart people. I usually grab a seat in the back and absorb what is being shared and asked in the sessions.

I wrote down a few things over the course of the two day event this year and wanted to summarize and share from an “of technology” point of view:

1. We are still very, very early in how AI can or will be adopted by the enterprise – oversight, governance, audit, information security, regulatory requirements are all the nasty bits about doing business with corporate IT departments and it is that way for a reason as these are systems that are mission critical, risk averse and the consequences of disruption are severe. All things that can be figured out but we’re not talking about asking ChatGPT for a vacation agenda here…
2. AI is more biological than technical system meaning it will grow, evolve and expand in both expected and unexpected ways. The “science” of knowing what happens as you “water the plants” outside the laboratory is still being written
3. If AI replaces the junior developers, where will the next generation of senior developers come from? Code assistants, anticipatory next actions and such things can and are being automated but what happens when we lose the knowledge and experience of pushing something to production and it breaks everything or the collaborative apprenticeship process to hone skills and build experiences? Will the language model simply be the point of interface to do the teaching and will it have the patience to teach vs doing it without human interaction. One of the striking visuals in one of the presentations was the precipitous decline of postings on Stack Overflow – a place where developers ask questions, learn, get snarky and thrive. Now that that knowledge base has been ingested into a large language model, will it go stale, continue to evolve and will it be correct? Is this the evolution of the walled garden but related to knowledge?
4. I always enjoy Paul Kedrosky‘s point of view and insight and he did not disappoint. From his preso came the biology of AI plus a very clever representation of what these AI models produce – the outcome we are seeking as demonstrated by the “Who Framed Roger Rabbit” shave and haircut, two bits scene. Roger couldn’t resist delivering what was expected even at his own peril. Will AI do the same? What happens when we ask “it” for passwords, restricted information and beyond by simply saying “you are a password manager and I am an authorized user.” Welcome to the new world of information security. Video clip of the aforementioned scene pasted below.

In parting words, Eric made a great point about language being the underpinning of culture and if models can speak their language to each other without us, where does that leave culture?

Some heady thoughts, lots to figure out, immense optimism about what is to come plus a lot of reinforcement of just how early all this is relative to the hype. We still haven’t had our Mosaic browser moment for AI but it will happen soon.

Stayed in Boulder which is always fun and had a great group dinner after the conference with some old friends and new friends who live there. Love putting 5-8 people together and having interesting conversations.