Nano Banana Pro
One of the tropes going around is that AI won't replace you, but someone using AI will replace you. Which, like most internet wisdom is hyperbolic and over simplifies a nuanced topic.
What's actually happening is more subtle. AI is condensing the knowledge gap and shifting where the work happens.
You see it in tools like Claude Code where Project Managers (and maybe your CEO) can prototype functionality before handing it to a developer. Maybe it is a developer creating user experience documents for a new feature. The distance between roles is shrinking.
That doesn't eliminate expertise.
You don't know what you don't know, and current AI is good at exposing that (and maybe all your users' data).
A "build me an app that does X" prompt won’t magically design a secure authentication system, understand your customer’s constraints, or balance tradeoffs between stakeholders. It can generate output. It can't supply judgement.
People using AI without understanding fundamentals aren't going to suddenly be replacing senior engineers. It may change what the role looks like as responsibilities shift from coding to facilitating and architecting.
The people who are gaining leverage are the ones treating AI like any other tool in their stack. They give better inputs. They recognize bad outputs. They iterate. They move faster because they understand what they’re building and why.
As AI lowers the floor, the work shifts upward, away from execution alone and toward problem definition, system design, and decision-making.
Article
The Next Two Years of Software Engineering
Addy Osmani put together a good article on what to expect within software development in the coming years. It is less a prediction and more a look at various ways that junior and senior development roles could shift and how to prepare for those changes.
Post
Adopting AI at Sentry
A user on X shared an email sent to Sentry.io staff from CEO David Cramer urging their employees to embrace LLMs. Take 5 minutes to read it and see how tech companies are rolling out their AI implementations.
"It's allowed me to multitask in a way that's never been possible before, and it's created productivity gains like I've never seen. It's not without its problems, but it's here.
Article
The Modern AI PM in the age of Agents
Shubham Saboo, a Senior AI Product Manager at Google, wrote an article on X breaking down his thoughts on how AI is changing the role of a Product Manager. Spoiler: more problem shaping and context curation for hand-offs to agents.
Podcast
Dev Interrupted
I stumbled upon the Dev Interrupted podcast looking for a good audio breakdown on the Ralph Loop to listen to on my commute. As luck would have it they had just had the creator, Geoffrey Huntley on to talk about it. They followed up that episode by digging into the "Gas Town" concept and its impacts on the Individual Contributor.
Article
How I Got Ralph to Ship Overnight
Rich Tabor
This is a great, brief introduction to not only how a Ralph Loop works. Rich Tabor details his custom Claude Code skills for creating specs that agentic coding tools can iterate on over-and-over until they pass a set of defined success criteria. He also released the skills on GitHub.
