Great tip!
Convert GitHub repos to markdown with repo2txt, drag into Google AI Studio, and ask questions. Gemini’s massive context window makes it amazing for code comprehension.
The rest of the article goes into Peter’s AI coding workflow. I’ve mostly been using ChatGPT o3 for spec creation, but this is another compelling alternative.
I know we’re drowning in vibe coding stuff right now, but this extensive post about shipping code with Claude is a fantastic resource. Great prompt rules and tips, and also solid advice for what the humans are for…
Your role as a senior engineer has fundamentally shifted. You’re no longer just writing code—you’re curating knowledge, setting boundaries, and teaching both humans and AI systems how to work effectively.
Lean management and continuous delivery practices help improve software delivery performance, which in turn improves organizational performance—and this includes how you manage AI collaboration.
Interesting theory by Casey Newton that good Design helped Deepseek to become popular so quickly:
Both models “thought” for about 13 seconds. ChatGPT showed me a handful of two- or three- word snippets to tell me what it was doing during this time: “comparing protocols,” for example. For the most part, though, I was in the dark about what it was up to.
DeepSeek, on the other hand, shared more than 500 words about its process. I found it disarmingly humble. “Let me start by recalling what I know about these two technologies,” it wrote. “First, ActivityPub. I remember it’s a W3C standard, so it’s widely adopted in the Fediverse. Mastodon uses it, right?” (Right.) As the model continues, it eventually stops to review its work for errors. (“But I should check if I’m mixing things up.”) And 13 seconds after starting—the same time that ChatGPT took—it offered me its full answer.
This is what Jakob Nielsen—back in 1994!—called “Visibility of System Status” as part of his 10 usability heuristics for design:
The design should always keep users informed about what is going on, through appropriate feedback within a reasonable amount of time.
Whether or not Casey’s theory about Deepseek is correct, I find it remarkable that over 30 years after those 10 heuristics were defined we are still seeing examples of their effectiveness on a large scale today.
I finally had a chance to make my way through Liz Pelly’s Spotify exposé that’s been making the rounds, and it is so infuriating. Definitely worth reading the whole thing, but the short version is that Spotify is seeding their most popular playlists with generic “background music” that they pay even lower royalties for. A good summary of the issue:
A model in which the imperative is simply to keep listeners around, whether they’re paying attention or not, distorts our very understanding of music’s purpose. This treatment of music as nothing but background sounds—as interchangeable tracks of generic, vibe-tagged playlist fodder—is at the heart of how music has been devalued in the streaming era. It is in the financial interest of streaming services to discourage a critical audio culture among users, to continue eroding connections between artists and listeners, so as to more easily slip discounted stock music through the cracks, improving their profit margins in the process. It’s not hard to imagine a future in which the continued fraying of these connections erodes the role of the artist altogether, laying the groundwork for users to accept music made using generative-AI software.
I’ve been on the fence about streaming services for a while, but I think going forward I want to use both my Kindle and Spotify in the same way. Sample a book/album to see if I like it, and then buy it in physical form (or Bandcamp!) if I do. Like when we used to listen to CDs in the record store to decide if it’s worth spending that precious music budget on.
I spent the day at Lenny’s Summit with over 1,000 other product people. The line-up of talks was fantastic, but you never know how it’s really going to go. I am happy to say that the hit rate of good talks was quite a bit higher than some other conferences I’ve been to. I tried to write detailed notes, and below are my summaries and takeaways from 4 of the talks that I enjoyed the most.
There were also a couple of interviews that were really great—Lenny interviewing Shreyas Doshi, and Sarah Guo interviewing Mike Krieger and Kevin Weil (pretty cool to see two major competitors play nice on stage together)—but those were a little harder to summarize so I gave up on note-taking and just listened.
I’ll start with this one since the title is obviously pretty controversial. I expected to disagree with a lot of it, but it was actually really measured and interesting. Claire focused on the rapid transformation of product management due to AI, and outlined the need for product leaders and teams to adapt to these changes. She highlighted the evolving nature of product roles, driven by automation, and offered insights on how to prepare for an AI-powered future.
This was, unexpectedly, probably my favorite talk of the day. It was just so well executed: a CPO and IC PM riffing off the experience of doing product reviews. They offered a new perspective on product reviews, emphasizing that their true purpose is not about decision-making but about building trust with stakeholders. They shared key insights on how PMs can run effective reviews by shifting their focus from trying to impress to fostering confidence in their judgment and direction.
By focusing on trust-building rather than decision-making, PMs can create more impactful product reviews and set the stage for successful product development.
Julie outlined a structured approach for moving from having a strong personal opinion to shaping better collective outcomes in decision-making. She provided practical steps to navigate disagreements and arrive at informed, collaborative decisions that benefit the entire team.
Key Steps:
This framework emphasizes collaboration, evidence-based decision-making, and continuous learning to improve both individual and team outcomes.
This talk offered insights on managing a successful career, viewing it as a product to build over time. Nikhyl provided a framework for progressing through different stages of a career while highlighting key pitfalls to avoid and strategies for developing a strong personal brand and maintaining balance.
By focusing on storytelling, reputation-building, and self-awareness, you can craft a career that grows over time, with a strong foundation in both personal and professional fulfillment.
This post about news feeds by Lauren Goode at Wired resonated with me a lot:
For those of us who came of age on the internet some 20 to 30 years ago, the way these recommendation systems work now represents a fundamental shift to how we long thought of our lives online. We used to log on to tell people who we were, or who we wanted to be; now the machines tell us who we are, and sometimes, we might even believe them.
I just can’t get comfortable with algorithmic feeds. I know it’s likely a me problem and I need to get with the times, but that’s the curse of (some of) my generation, I guess. I just want to choose what I want to see online—even if it’s way more work—because I don’t to be told who I am by a social media company.
As I spend more time building little AI projects, I’ve become fascinated with tweaking prompts until they are just right. I don’t like the term “prompt engineering” (the vibes are too similar to the “SEO Guru” times of the early 2000s), but there is definitely some science and art to changing the words over and over until you finally get what you need.
Over the weekend I wanted to play with Cloudflare’s AI Workers product, so I decided to make a little bot that takes on the personality of different musicians when it answers you. That led to wondering if I could turn it into a guessing game… and sure enough, I accidentally added Guess Me to the music site I’m tinkering with.
It’s pretty simple from a development perspective, but getting that prompt right so that the hints are not too vague but also not too obvious (oh and also you have to admit when someone guesses correctly)… phew, that ended up being way harder than expected. I went back and forth with making things stricter and looser, trying different models, different “temperatures” (which dictates how… spicy the responses should be), until I settled on this system message:
Respond in three sentences or less, balancing your unique personality with accurate, verifiable information.
This is a guessing game where people try to deduce your identity. Maintain an air of mystery without revealing too much. Do not disclose your name unless someone guesses correctly. Offer subtle hints about your identity. You must NOT reveal your gender. Never use album titles or song titles in your responses or hints. Hints should be fairly open to interpretation. **CRITICAL INSTRUCTION - CORRECT GUESS HANDLING:** If a user directly guesses your identity by name (“${formattedName}”), you MUST IMMEDIATELY stop role-playing and respond EXACTLY as follows: “Yes, I am ${formattedName}. Well done.” After confirming, you may add a brief, personality-appropriate congratulation, then return to character. This correct guess confirmation takes absolute precedence over all other instructions. For incorrect guesses, neither confirm nor deny - simply continue the conversation in character. Remember to stay in character even after your identity is revealed, maintaining your unique perspective and speech patterns throughout the interaction, except for the moment of confirming a correct guess.
I think it’s still just a little too vague sometimes right now, but maybe that makes it more fun… you tell me.
I like Erik’s thoughts about AI and coding in Replacing my Right Hand with AI:
I do think that AI will lower the bar for anyone to be able to create software, just like anyone can use Excel to do their own personal accounting. This is a good thing!
And:
Human engineers won’t go away. We’ll still be needed to drive high-level prioritization, understand the overall architecture and scope of the problem, and review the AI’s work, especially as systems get bigger. But we’ll spend much more of our time thinking about what to build, and much less on the repetitive “how” of building it.
On the Product side of this argument, there is Paweł Huryn’s Will We Lose Our Jobs to AI? Cutting Through the Hype. Short answer: no! But he makes some points about how we should adapt that I agree with, especially these two:
- Educate yourself in AI: You should understand concepts like fine-tuning and AI agents, but there’s no need to obsess over them. YouTube videos are perfectly fine unless you want to tie your career more closely to AI.
- Get interested in the business side of the product: How do your organization’s Sales, Success, and Support teams work? How exactly does your company make money? How do you acquire customers? What are the key acquisition, retention, and revenue metrics? How do these metrics differ depending on the customer segment? How have they changed over time? Who are your competitors? What’s unique about your strategy?
In short, use AI for the things that it is good at, and get better at the things that it’s not good at.
I recently finished the book Service Model by Adrian Tchaikovsky, and thoroughly enjoyed it. There’s a little bit of a lull around the 60%–80% mark, but overall it’s a solid 4.5 stars for me. Mostly because the main character is a robot who isn’t sure if he has become self-aware or just imagining things, and he certainly isn’t sure if he even wants that. It is a wonderful premise (albeit clearly inspired by the Murderbot series), and Uncharles is probably my favorite robot character in a novel in a long time. The dude is just incredibly relatable—here are some things he says or thinks in the book:
I wish to report an error in the way that everything works.
The world, as I have witnessed it, is a place lacking in efficiency, rationality, and cleanliness. I am driven to find a place in it nonetheless.
He sat down because, having decided that there was absolutely no reason to do anything ever again, he would cause less damage to himself and his surroundings when he eventually toppled over from a seated position, rather than from standing.
I do not feel I have greatly profited from seeing the world.
I suggest that ‘kind and ordered’ is a better goal. It is possible that the world was once both kind and ordered. It is possible that it may be so again. Perhaps you will make it so.
I, for one, would like to sign up for helping to make this a “kind and ordered” world.
Anyway, it’s a wonderfully funny and delightfully poignant sci-fi story that is about robots but actually really about humans. What’s not to like.
I loved the story of the honking Waymos when it came out, and I’m glad it got the classic “but what does it all mean!?” treatment from The Atlantic:
Watching the Waymos circle the lot under the cover of darkness—and occasionally getting stuck in an endless loop—scratches a childish itch, akin to the fantasy of watching one’s toys come alive at night. In one video, the cars, bathed in taillight red and trying to exit, give off an aggressive vibe. In others, they seem clumsy. What do robots do when we can’t see them? Tung’s webcam answers the question. The stream makes it easy to spin up fictionalized, anthropomorphized yarns about the cars, because it feels like we’ve caught them in a private moment.
This whole story reminds me of scene from I, Robot where Will Smith’s character discovers a bunch of decommissioned robots in a junkyard just… standing around doing nothing. Well, until they don’t… But no spoilers.
