Michael Werle

I made an explorable esoteric research environment with a Custodian named MERIDIAN that can launch its own agent swarms to solve problems. Does isopsephic arithmetic and lookup for any Neoplatonists :-) https://lnkd.in/gYzqVn7m

1

1 Comment

Anthropic built Cowork in 10 days—largely written by Claude Code itself. The team said they "spent more time on product and architecture decisions than writing individual lines of code." This is the new shape of software engineering. The bottleneck isn't typing. It's taste: knowing what to build, how to structure it, when to say no. Code generation is table stakes now. System design is the moat. https://lnkd.in/gij_qHgU

7

4 Comments

I thought this was a good and very reasonable talk by Eleanor Millman on how to prioritize projects for a platform engineering team: https://lnkd.in/gaYtUHRH A few insights I took from it: - Unlike product features which can be more directly tied to revenue, platform engineering work is a couple steps removed. But that doesn't mean it's unimportant! Far from it, platform engineering projects can make everyone at the company more efficient, able to produce higher quality software, etc. - While "urgency" is always going to be a factor, you don't want to just be putting out fires, you want be able to prioritize work which is "high impact" ... even if that impact isn't felt immediately. - Good rule of thumb: prioritize "highest impact for lowest effort". - The talk has some ideas on what factors to choose for impact, and how to blend them. For instance "speed of development" is one factor, "cloud cost optimization" might be another. The weighting of different impact factors can change over time, depending on the needs of the business. I would say that a lot of companies don't have much methodology here but I can really see the value in codifying it. You can always change the methodology or the weighting if it's spitting out results that don't pass the smell test or you really feel it is leading the company astray. Clarity can give the org more freedom to put resources behind projects that would otherwise never be taken on. When things are unclear, prioritization still happens implicitly, but the decisions tend to be a lot more random and fear-based and the org is worse off as a result.

15

3 Comments

I had a crazy idea: what if recommenders stopped being “similarity engines” and became relationship engines? GraphRAG + rec systems sounds wild… until you try it on real estate: commute • HOA rules • school zones • neighborhood vibe • “quiet but walkable” preferences I wrote a breakdown on when GraphRAG helps, how to build it without overengineering, and when it’s a complete waste of time. Read it here 👇 🔗 https://lnkd.in/gqfUqrFJ #GraphRAG #RAG #RecommendationSystems #AppliedAI #LLM #MachineLearning #AIEngineering

32

1 Comment

I was reading through some of the comments on another post of mine, and one from Randy Clark caught my eye. He said, "I think what you’re describing...feels less like an 'AI slop' issue and more like an ownership gap...." Essentially, "AI slop is an ownership gap." That line sat with me for a while. Not because it was provocative, but because he was right. It made a lot of sense, and so I wanted to expand on it here in this post. Here are the ownership gaps I came up with. 🎯 The accountability gap When AI-generated code ships, responsibility gets fuzzy. No clear author for correctness. Tool makers say, "we just provide the model," users say, "I just used the assistant." Quality improves when someone still owns the final outcome. 🚪 The curation gap AI increases output, but judgment still decides what belongs. Fewer reviews before merge, more copy-paste without understanding. This is where slop slips through. Review is not friction; it is stewardship. 💰 The economic value gap Volume looks like progress until the costs land later. More output feels productive, but short-term shipping often beats maintainability. AI slop creates technical debt that lands on teams who didn't choose it. Sustainable value requires balancing velocity with durability. 🎨 The cultural stewardship gap Craft matters, even when machines assist creation. Less pride in readability, fewer thoughtful docs, more "works on my machine" code. When code is treated as disposable, slop becomes the default. Pride in systems signals care, not perfectionism. 📚 The training and skills gap AI helps most when humans keep sharpening judgment. Models train on generated code, and developers rely on AI instead of building depth. Fewer experts can spot subtle issues. Slop feeds on itself: worse training data plus weaker human judgment. Depth enables people to recognize good versus acceptable. ⚖️ The incentive alignment gap People follow what metrics reward, intentionally or not. Velocity over sustainability, individual gains diverge from collective outcomes. Healthy incentives protect quality without slowing innovation. None of these gaps imply failure or neglect. They are natural pressures in fast-moving AI environments. Randy's comment helped me see the pattern more clearly. AI slop thrives where ownership dissolves. Ownership is the quiet force behind trustworthy AI. When ownership is clear, quality has a fighting chance. Do you also view AI slop as an ownership gap? #AISlop #AIOwnership #AIGaps #AIAccountability

26

16 Comments