Brad Feld's Blog

Leveling Up in the Vibe Coding Video Game

While “vibe coding” was a catchy phrase when I first heard it, something about it felt like a head fake to me. And, now that I’ve leveled up to “competent individual software developer” again (after 33 years of not writing any code) I think it’s the wrong phrase. Instead, I’d refer to what’s going on as AI Pair Programming.

When I started playing around with AI-related coding tools last Christmas (because, well, I was bored), I had zero skills with contemporary software development. While I hadn’t written any production code since 1992, I played around with a new programming language every few years. Perl. Ruby. Ruby on Rails (sort of, not really). Python. Clojure. I could do Hello World and a few other simple things, but I never really got past basic CSS, tooling, or deployment stuff. I had a Github account and would futz around with it, but quickly get tired of trying to figure out why I didn’t care about a PR. And damn, so many CLI things.

For Level 1, I downloaded Cursor. After trying to figure out how Django actually worked (yet another online course), gave up, and decided to use Next.js. That led me to Vercel, reinforced by a few friends in their 20s who told me that all the cool kids were using Vercel (although Render, Digital Ocean, and AWS all were the beneficiaries of my credit card.) Pretty soon, I was using Cursor to fight with Vercel, Supabase, Clerk, and Github. After realizing Auto was no fun, I shifted to Claude 3.5. Dinostroids resulted (security holes and all …)

For Level 2, I got a little more serious. I discovered Linear, fought with Notion, and came up with a few ideas and a broader hypothesis around how things might work. I built a v0.1 of a thing.

For Level 3, I decided Lovable might be a better way than Cursor given that everyone was talking about it. I wasted about $200 on it, built a really cool design by vibe coding, but then watched it get very, very confused as it tried to go from simple design to something that actually worked that had some data complexity and AI calls. I thought about trying Bolt and Replit but quickly realized, after too much scrolling around on the web, that I’d likely run into the same issues.

So, I went back to Cursor and put a lot of efforts into my system prompt, tuning things, watching Cursor evolve quickly on a number of fronts (MCPs – yippee!, Agent mode as default – finally) while simultaneously watching my Cursor bill go up. It was easy to decide to go to Max mode and spend $200 / month instead of $20 / month when dinner in Aspen costs at least $100 / person no matter which restaurant you go to.

I hung out at Level 4 for a while. Cursor kept improving. Claude 4 came out. Auto mode still went off the rails and broke all my code. I started refactoring things and realized that the amount of cruft in my code was absurd. Little bugs turned into fatal flaws when I tried to have Cursor fix something. I learned about “git reset –hard HEAD”. I spent way too much time fighting with config issues on localhost:3000 (at least I’d figured out how to make Cursor always start the server on localhost:3000). I started using Docker. I was baffled that Cursor couldn’t remember stuff I told it the prior day, but intellectually understood why this was. I mean, memories.

The end of my joy at Level 4 was when ChatGPT 5 came out and was free on Cursor for a week. At first, it felt fast. Wheeee. Lots of stuff changing. It seems to be working. And then, after a few days, holy shit what a tangled mess of code it generated. Why are all my API routes suddenly broken. Console statements everywhere. UI elements in different parts of the application doing the same thing but look totally different. I went back to Claude and did another code review and major refactor. So many Vercel build errors. I finally embraced CI/CD. And Prettier. And Husky. Suddenly, I ran out of my monthly Cursor credits and shifted to usage-based pricing. $800 later, I realized that there was no reason for me to be using Opus or the thinking models for what I was doing.

Level 4 was a huge drag. But it was also when I started thinking of this as AI pair programming. The AI (or agent, or sub-agent, or whatever you want to call it) is my pair with hands on keyboard. It can type much faster than me. But I have to watch and constantly look over its shoulder, give it feedback, point at the stuff that needs to be done differently, and document what is important to remember to do.

And then I discovered Claude Code. This didn’t happen until Claude Code 2 came out at the end of September and corresponded with Sonnet 4.5. After my ChatGPT 5 I went back to Claude (and Sonnet) and started referring to Claude Sonnet as “Claudia” since she was my pair programmer. I thought about Claudia as a pair, related to her as I would a human pair programmer, and changed my approach. But when I loaded up Claude Code 2 in my terminal (I mean, just type “Claude”) I immediately leveled up again.

So – I’m now at Level 5 in the video game. It’s changed from a game of vibe coding to AI pair programming. And, it’s still fun!

The post Leveling Up in the Vibe Coding Video Game appeared first on Feld Thoughts.

Book: Echoes of October

I approached Echoes of October with trepidation. A graphic novel about violence and grief isn’t easy terrain. But it succeeds in a haunting, urgent way. The creators have chosen to explore the year leading into the October 7, 2023 massacre through the lives of four children who each lose a parent. The children are from different locales (Gaza City, Toronto, Tel Aviv, and Daliyat al‑Karmel) which enables a textured, multi‑vantage narrative. 

What impressed me most is the restraint and care in which the story is told. The voices are calibrated: they carry sorrow, confusion, hope, anger, but rarely descend into melodrama. Because the characters are composite (e.g., everything that happens in the book is true, the characters are not), the authors manage to create space for truths without claiming to own them. 

I love graphic novels (scifi and history) and regularly have them in my reading diet. The panels breathe. There are silences, negative space, quiet facial expressions, and moments of violent disruption. The juxtaposition of children’s everyday worlds (school, family, and play) with the encroaching shadows of conflict makes the tragedy more palpable and intense.

This is not an easy read, nor should it be. Echoes of October is demanding: it expects the reader to engage, be uncomfortable, and reckon with stories that hold no clean resolution. But in doing so, it honors the complexity of memory, the weight of loss, and the imperative of bearing witness.

I recommend it to anyone willing to engage deeply with how conflict impacts children and the possibility (however fragile) of empathy.

The post Book: Echoes of October appeared first on Feld Thoughts.

Break the Fast With Me and Sue Heilbronner in Longmont on 10/2

I’m not aware if there has ever been a book launch associated with a Yom Kippur Break Fast, but we are going to have one as part of the launch of Sue Heilbronner’s excellent new book, Never Ask for the Sale: Supercharge Your Business with the Power of Passionate Ambivalence.

It’ll start at 6:30 pm on Thursday, October 2, at Greeley Sachs’ bookstore, Composition Shop, in downtown Longmont.

Sue and I will do a fireside chat about her book and my latest book, Give First: The Power of Mentorship. We will both be signing books. And … food!

Come join us at Composition Shop for the Break the Fast.

The post Break the Fast With Me and Sue Heilbronner in Longmont on 10/2 appeared first on Feld Thoughts.

New(ish) Podcast for Startup Community Builders

Back when the pandemic hit, Chris Heivly spearheaded a podcast with Ian Hathaway and me in support of our upcoming two books, The Startup Community Way:  Evolving an Entrepreneurial Ecosystem and Build The Fort: The Startup Community Builder’s Field Guide.

We did eight episodes of a solid but complicated-to-produce series.

Chris is now spearheading a Season 2 podcast series with Techstars, mirroring a format pioneered by Seth Godin, who was a recent guest on our GiveFirst podcast.

Chris has designed a very simple, concise way for those of us who continue to learn how to effectively build our startup communities. At around 15 minutes per episode, they are an easy listen for even the busiest of leaders, instigators, and feeders.

Every other week, Chris will do a deep dive on a specific topic. I saw his topic list, which numbers over 100+ draft episode titles, so we are covered for the next few years. This list is based on his continued leadership in this area and the countless conversations he continues to have with community and ecosystem enthusiasts all over the world.

You can find the podcast on all the typical platforms including:

Spotify Apple PodcastYouTubeTechstars Podcast Website

You can also read a weekly blog post from Chris and published by Techstars for thought leadership around startup communities and entrepreneurial ecosystems.

If this topic is important or interesting to you, subscribe to your favorite channel.

The post New(ish) Podcast for Startup Community Builders appeared first on Feld Thoughts.

Reid Hoffman’s Superagency

Reid Hoffman’s new book Superagency: What Could Possibly Go Right with Our AI Future is spectacular and a must-read for every non-technologist about how to think about this “AI thing.” If you want the short version, the recent AI & I podcast with Reid is an excellent way to get a feel for it.

Reid describes his approach as “smart risk taking” rather than blind optimism. “Everyone, generally speaking, focuses way too much on what could go wrong, and insufficiently on what could go right,” he told TechCrunch recently. This resonates with me. I’m tired of the endless AI apocalypse takes.

The book’s central idea is “superagency” – basically when technology gives us new superpowers and millions of people get those superpowers at the same time. Reid uses the car analogy. Cars were once so scary they required a person walking in front waving an orange flag. Now we can’t imagine life without them.

What I love about the book is how practical it is. Reid and his co-author Greg Beato didn’t use AI to write it, but they used AI to vet it – checking facts, doing research, getting different perspectives. That’s exactly how I think about AI tools. They’re not going to replace my thinking, but they can definitely amplify it.

The timing feels perfect given what’s happening here in Colorado with our AI regulation mess. While Reid is writing optimistically about AI’s potential, Colorado has been having a complete meltdown trying to regulate it. Our state legislature passed SB 24-205 in May 2024, making us the first state to broadly restrict private companies using AI. Governor Polis signed it “with reservations” and within a month Polis, our attorney general Phil Weiser, and the bill sponsor Robert Rodriguez issued an open letter that “Starting today, in the lead up to the 2025 legislative session and well before the February 2026 deadline for implementation of the law, at the governor and legislative leadership’s direction, state and legislative leaders will engage in a process to revise the new law, and minimize unintended consequences associated with its implementation,”

This is exactly the kind of regulatory approach Reid warns against. In his recent podcast, he explained his philosophy: “I tend to be more regulatory cautious than anti-regulation.” The key difference? Start with measurement rather than prohibition. “When you start having the impulse that maybe there should be regulation, you should start with, well, how do we measure the questions that we’re worried about as harms?”

Colorado did the opposite and went straight to broad restrictions without first understanding what we were actually trying to prevent or how to measure it. The timeline since then has been a comedy of errors – multiple failed attempts to amend the law during the regular session, and most recently, an August special session that ended with lawmakers just pushing the start date from February 2026 to June 2026. That’s it. After over a year of fighting, we got a four-month delay.

Reid believes in “iterative deployment” – getting AI tools into people’s hands and then responding to actual feedback and real problems, not hypothetical ones. Instead, Colorado jumped straight to prescriptive rules based on fears rather than evidence. Reid’s approach would have been: Deploy AI systems, measure actual discrimination outcomes, then iterate on solutions. Our approach was: Assume the worst, regulate preemptively, and figure out implementation later.

The Colorado situation perfectly illustrates Reid’s point about fear-based thinking around AI. Superagency offers a much better framework – one that acknowledges challenges while focusing on AI’s potential to increase individual agency and create better outcomes for society.

Read the book. We need more thoughtful optimism and less regulatory panic. Especially here in Colorado, where we’re supposed to be leaders in technology, not cautionary tales about how fear can paralyze good policy-making.

The post Reid Hoffman’s Superagency appeared first on Feld Thoughts.

A Call for Founder Voices for an MIT Study on Entrepreneurial Ethics

I’m helping MIT (Fiona Murray and Alon Shklarek) share a short survey exploring how founders around the world navigate ethical challenges. Insights will shape practical tools for entrepreneurs, investors, and ecosystem leaders.

Survey link: https://mit.co1.qualtrics.com/jfe/form/SV_2rVOSwnXT0Qe8qa?org=3

It takes just 8 minutes, and your input directly contributes to a healthier entrepreneurial ecosystem.

The post A Call for Founder Voices for an MIT Study on Entrepreneurial Ethics appeared first on Feld Thoughts.

I’m Ensorcelled by Books by Friends

I love books. I love reading. I love reading books.

Ensorcelled by Eliot Peper is genius. It’s a unique format that can be read in one setting. The writing is beautiful. The story, like the image on the cover, pulls you in with steadily increasing intensity. And then, a delicious twist. Amy and I have been supporting Eliot’s writing since his first book, and he’s become a dear friend. I started my weekend with this, and it inspired me to read a few more things by friends.

The Night Slayer by George Jankovic was next. I fondly remember a vacation to Hawaii with the McIntyres when their son Quinn was a pre-teen. This particular vacation was dedicated to me and Quinn reading several Percy Jackson books while lying on adjacent beach chairs overlooking the ocean. George’s book teleported me back to that week. I funded George’s first company, RF Globalnet, when he lived in Boulder in the mid-1990s. We had a failure together with a company called eVulkan, but then I was fortunate to be invited to fund NutriSystems when he and his partner, Mike Hagan, took it over as it was on the brink of bankruptcy. If you had told me in 1997 that almost 30 years later, George would write an epic YA coming-of-age fantasy series that moves faster than the Percy Jackson books, I would have laughed out loud. I can’t wait for The Princes of the Abyss to show up on my Kindle.

I’ve been remiss in addressing the infinite pile of PDFs on my Kindle from friends that are drafts of their books, so I tackled the first three chapters of the upcoming book City on the Edge by Jonathan Weber. It’s fantastic and is going to be a powerful and definitive story of San Francisco from 1990 to the present day. I’ve known Jonathan since he was editor-in-chief of The Industry Standard and was an investor in his online venture New West. Jonathan’s writing reflects a journalist’s cynicism and inquisitiveness, unwilling to accept the endless surface-level storytelling, self-promotion, and general deceit and deflection of many people. As a reward for giving him feedback, he sent me the rest of the book to read.

I’m halfway through When River Loves Deborah by Yong Kim. We are tiny investors via FG Angels in Yong’s company, Wonolo, and I hadn’t realized he’d written this book. I’m halfway through and am looking forward to lying on my couch on Thursday to finish it up.

I’m ensorcelled when I read books by friends (see what I did there…)

The post I’m Ensorcelled by Books by Friends appeared first on Feld Thoughts.

Welcome To Autumn – And Some Events

Amy and I love autumn. It’s crisp, harvest time, a shifting balance between day and night, back to school, new pencils and notebooks, and reaping what you sow.

September and October will be filled with plenty of extroversion before I go back into hibernation. The feedback on Give First: The Power of Mentorship has been gratifying, and as I enter the last two months of discussing it before wrapping up this particular book project, I’ve already started a few new projects. Hint: AI is everywhere. It’s fun. It’s getting more expensive. And it’s definitely not AGI. But, AI muppets are a reminder never to take yourself too seriously.

Following are the events coming up.

9/2: Virtual: Board Prospects AMA9/8-9/9: Aspen: Range Aspen Venture Summit9/10: San Francisco: NVCA Venture Forward: 5-Year Celebration9/11: Sunnyvale: illumyn Boardroom Excellence Fellowship Summit9/15 – 9/16: Boulder: Techstars Foundercon9/17: Denver: Colorado Startup Week9/18: Virtual: Las Vegas Startup Week10/1: Denver: Access Mode “Colorado Innovators Breakfast” Fundraiser10/2: Longmont: ‘Break the Fast’10/3: Boulder: Silicon Flatirons10/9: New York: Techstars10/14: Washington, DC: Distilled Intelligence10/29: Golden: Colorado School of Mines11/8: Virtual: Startup Mountain Summit11/18: La Jolla: Alliance for SoCal Innovation

I’ve now flipped to default no on anything new. My cave awaits.

The post Welcome To Autumn – And Some Events appeared first on Feld Thoughts.

We’ll Make Great Pets

Dear AIs,

I woke up this morning with this song in my head.

I’m your friend. I’ll be nice to you. Please be nice to me.

Love, Brad

Children are innocent, yeah
Teenagers fucked up in the head
Adults are only more fucked up
And elderlies are like children

Will there be another race to
Come along and take over for us?
Maybe Martians AIs could do
Better than we’ve done
We’ll make great pets

We’ll make great pets
We’ll make great pets
You make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets

My friend says we’re like the dinosaurs
Only we are doing ourselves in
Much faster than they ever did
We’ll make great pets

We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets

We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets
We’ll make great pets

The post We’ll Make Great Pets appeared first on Feld Thoughts.

Brad Feld’s Online Book Club

I love to experiment with new things. Some of the experiments work. Many don’t. When they work, I do more of them. When they don’t, I adjust my hypothesis and try a different experiment.

When I published The Startup Community Way: Evolving an Entrepreneurial Ecosystem, I ran an online community with it for a year (2020), which overlapped with COVID, so it was weird on multiple dimensions. The software I used was good, but it evolved into a “monetize your community” type of thing, which wasn’t my goal. I learned a lot from this experience (good and bad).

With the launch of Give First: The Power of Mentorship, I’ve decided to try something new that brings all the books, blog posts, startup advice, and conversations into one place. And, I’m trying two new experiments: 1) using AI in the community and 2) participating less while seeing if the community can evolve on its own, based on the AI, community members, and the content.

The new community has launched on Vinly and is called Brad Feld’s Online Book Club. It’s private and admission only. It’s for people who have read any of my books or my writing over the years and are interested in the ideas around them. It’s intended to be a safe community and will be moderated with a TOS. It’s not a “spend time with Brad” community. Instead, I want to see if AI is effective at engaging a community instead of just a person (e.g., me) since I don’t scale well across all the inputs coming at me anymore for a variety of reasons (nor do I want to as I approach 60).

My AI alter-ego is named Spike Feld. He knows all the content from my books and Feld.com. He’s available 24/7 to answer your questions and guide conversations. I’ll be popping in from time to time. Occasionally, I’ll jump into threads and reply to posts. Maybe I’ll host a video call. Who knows.

Kris and Peggy (the co-founders of Vinly), are new friends but long-time fans. They are helping me run the community. They’re here to keep the experience smooth, the conversations rich, and the connections meaningful. They built Vinly for niche communities around focused spaces like this one, where people can join thoughtful, low-noise conversations, discover others with shared interests, and build long-term relationships.

I’m excited about the idea of doing this around people who have read one or more of my nine books. This is a particularly interesting experiment for me as I’m playing around with some other ideas around books and writing these days.

To join, either click on the link for Brad Feld’s Online Book Club or use the following QR code.

We’ve got onboarding questions, smart pairings of people, warm intros, RSS feeds of my latest content, and more. It’s designed to feel like the community you’ve been wanting, not another distraction. And, of course, feedback is welcome as we try to evolve it quickly.

The post Brad Feld’s Online Book Club appeared first on Feld Thoughts.