At the end of the day, you gotta still use your brain. At this point, the model is a reflection of your brain.
CJ Hess retweeted
WTF is a smoke test?
This is a question software developers working with coding agents have probably asked themselves at least once.
The whole concept of "smoke testing" has become an inside joke, as it is usually a clear indicator of AI slop code.
Things like “synthetic smoke tests” or "Canonical UI smokes", are dead giveaways.
My coworker, @romanvellante, and I kept coming back to the same question though:
Why do smoke tests suddenly seem to be everywhere?
So we dug into it, and wrote this article covering:
1. Where smoke tests came from
2. Why coding agents keep creating them
3. When they are useful, and when they become clutter
4. How to keep repositories that coding agents work in maintainable
Hope everyone finds this piece helpful!
I understand the mentality of some of these comments. The models are gobbling the digital world but the real world is much harder and there are many input and output problems left to solve.
That being said, I think with enough tokens, this will work. Over the next few weeks, I am going to try and have claude manage the quote -> manufacturing process and if I can build this. I'm sure some changes will be needed but I agree, I want the physical thing.
We are releasing → classcad.ai today: AI driven embeddable CAD. If you ever enjoyed my OSS projects (zustand, react-spring, r3f, pmndrs), these all came out of CC. 🍻
A small 🇨🇭 team has been working on CAD for so many years, release seemed always out of reach. This is a huge milestone for us and i'd appreciate a RT. 🙏
My buddy who works in CAD said AI isn't good enough for him yet.
I took that as a challenge and wanted to prove to him this is a limitation of CAD tools, not the models.
3 beers later, we decided to try something out. Personal CAD software that builds CAD models.
I'll let you be the judge...
First prompt (sorry, this is what good prompting looks like):
"""
I'm going to have someone describe to you a 3D object on the fly in 3JS. Make a 3D CAD rendering thing that can export to an STL file, and maybe break it up into STL files or something for 3D printing. I don't know. Just imagine that it's 3D design software, and we need to have some controls to expand it and put together the assembly and a bunch of shit like that. Imagine you're a good CAD engineer, and it's got everything you need.
You're also going to design the thing that this guy's about to tell you. He's about to tell you what you're going to make now. Oh God, all right, you're going to make a three-axis metal 3D printer. It's got to be super fucking detailed, unbelievably detailed. I want you to run through this and then use Opus 55 Max sub-agents to review different categories here. Go online and understand what the Gauntlet loop is. I want you to run that with super, super critical reviewers until this is utterly perfect, and then open it in the browser.
"""
Shoutout @mattshumer_
"""
Second prompt just to get the video:
"""
Can you add a button that does a few things
1 - I want a floaty cursor style demo to play that shows all the features of the CAD software
2 - I want it to show off what was built and flash some full screen metrics on the actualy build and desing
I will then press that button and screen record the flow
"""
If you're in NYC next Thursday, come hang out! luma.com/2o2yus8t?tk=wPD0iE
CJ Hess retweeted
We are launching AI Agent Bootcamps at @tenex_labs!
What's an agent bootcamp?
A 5-day, free intensive to find & build high-impact agent opportunities with your real systems & data.
We've helped dozens of midmarket cos, F2000 enterprises, and PE portcos deploy agents into production. With bootcamps, we're lowering the friction of experiencing the power of an agent within your company's workflows to ~zero.
3 phases of the bootcamp:
1) Find Opportunities - We interview the people closest to the work to identify where agents could have the biggest impact.
2) Map Workflows - We map workflows step by step, including the systems involved and expected business impact.
3) Prototype Agents - We build prototypes on your data, demonstrate the results, and recommend what to take live first.
Bootcamps are led by our world-class FDEs & strategists, on the ground, working with your teams and systems.
What do you get?
1) A clear AI roadmap with ROI per opportunity
2) Up to 8 mapped opportunities
3) Up to 4 working prototypes
Who are agent bootcamps for & how much do they cost?
Midmarket & enterprise companies doing >$100m in revenue or funding. Bootcamps are free. They're an opportunity to see what AI agents can do for you before committing.
Why are we launching this now?
Many companies are in the messy middle of AI transformation. They've rolled out LLMs companywide, perhaps done some AI enablement work, coding agents are being used widely in engineering, and that's it.
The next obvious step in the AI journey is bringing agents to the workflows of an organization, but most companies are getting stuck.
Taking agents to production is complex and nuanced and we've built a repeatable process for handling the final mile of applied AI, which allows us to take agents into production in ~2 weeks.
How do I sign up?
Join the Agent Bootcamp waitlist here: form.typeform.com/to/vk5i1x9…
CJ Hess retweeted
The year after selling my business was one of the hardest periods of my life.
People are usually shocked to hear this & can't make it make sense.
"You accomplished what every entrepreneur dreams of."
"You made life changing money, what could be so bad?"
"You were given the gift of time, which is the greatest gift in the world."
What you're wondering is true & valid...and that's partially what made it so painful.
The journey of exiting @MorningBrew was a paradox of the highest order.
The outcome was beautiful in so many ways...
- A small idea that my cofounder and I dreamt in our dorm rooms at college blossomed into a media company that impacted millions of lives & reached a size that pushed our wildest imagination and ambition.
- A promise fulfilled. My dad passed away when I was a junior in college. In my last conversation with him, I promised him that I'd do everything in my power to take care of the Lieberman household so no on had to worry about money.
- Financial security that left me & my wife with a sense of security and optionality as we looked forward to growing our family.
- A new home for our business that could allow us to dream bigger & take bigger swings with the opportunity for greater scale, imagination, and impact.
It was also painful in nearly as many ways...
- My company ate my identity. I spent most of my young adult years building this business and being validated as "CEO of Morning Brew" which felt good, appealed to my ego, and continued this cycle of have an idea --> do good work --> get celebrated for it.
- My 12/10 motivation disappeared overnight. Fear-based motivators were the story of what drove me during this chapter of my life. I had a chip on my shoulder to prove those who bullied me (earlier in life) wrong & I felt financial anxiety to create safety for my family after my dad's death. Once we sold, the motivators that served me in the previous 10 years would not serve me for the next 10.
- The aftermath was lonely. My "struggles" post-exit felt like a champagne problem people would kill for. Because of that I rarely shared the journey. And because I rarely shared the journey, I bottled up many of these emotions.
- Time is a gift and a curse. Once I stepped out of the CEO role, I was given the "gift" of time. But unless you know how to productively direct time, it can feel like a prison. For a year straight, I marinated in my thoughts around being lost, no longer being motivated, and needing to re-find purpose.
While the 1-2 years post-exit were challenging, I now very much view it as a period of getting lost to better find myself.
I emerged from this chapter of my life with clarity around a few things:
1) Fear-based motivation is incredibly powerful but short-lived. Once you've nurtured your ego, materialistic needs, and childhood traumas, the fire under your ass quickly turns into fizzling coals.
Today, I try my best to listen to growth-based motivators that feed 3-year-old Alex's desires: curiosity, growth and play. This type of motivation is the energy equivalent of an ultramarathon vs. a 100-yd dash.
2) Freedom of time ≠ happiness. It is true that time is finite & having the privilege of choosing how you spend your time is exactly that...a privilege. But being gifted with more time does not mean you'll automatically feel more fulfilled. Why?
Our job as humans is to create. To create families. To create communities. To create companies. When we are not in the act of creation, or at the very least have clarity around what we're working towards creating, we can feel lost.
3) Deeply understanding your zone of genius & your values is the gift that keeps on giving. During my post-exit journey I got clear on two things:
- The 1-2 superpowers I bring to the world & how they can show up in my professional & personal life. For me: those are relentless curiosity & a fervor for storytelling.
- How money can be translated into happiness & at what point more money doesn't convert into more happiness. Money feels most valuable to me when it's used in 3 ways: not having to worry about it, giving to my family, and spending on experiences with the people i love.
The most helpful tool that helped me get clear is something called The Best Stuff Exercise (by Conscious Leadership Group), which I couldn't recommend enough.
This was a long-winded way of saying that everyone's journey has nuance and periods of wandering are uniquely human, profound & beautiful, but also complicated and painful.
Hope this helps!
Opus 5.5 and Astra have me rethinking the baseline again. A few thoughts that I feel like went from an idea to concrete in my mind over the past few months:
1/ Delete tests - the models love to slop up tests and check that “abc” == “abc” all over the place. A lot of tests restate the code instead of test the code. This coupled with the fact that the models just write correct code these days has me feeling like unit tests are now just a waste of tokens.
2/ PRs and code review are a ceremony at this point - maybe on super critical pieces like payments, auth, etc. it’s still worth going line by line. Otherwise, have the model write the code, have the model test the code, have models review the code, and have the model drive computer use to prove that it works end to end. Stolen from @thorstenball , “give me irrefutable proof this works end to end” often I ask for this in the form of a video and then I review demos and poke around myself.
3/ It’s orbing time - Of course everything is moving to cloud. I have discussed @AmpCode a lot. You can see this pattern with Codex cloud from dev day. Moving your work to the cloud is a second unlock that is not about using the models well.
4/ The bar is higher - Motion design videos just got commoditized. You need to innovate where the models struggle (not much room anymore) or else you’re ngmi.
5/ Teammate, not tool - The models these days can triage and monitor. I don’t push releases without a thread that’s watching it, reading logs, and writing some doc. At worst case, I build up context about what actually happens in production. Best case, I get a hot fix automatically.
6/ Hillclimb - Overused word rn, but performance, optimization, and sweeping refactors are no longer scary. If you close the loop on task → validation and it can be proven on a computer, it’s solved.
I think some of us were early to jump to these conclusions with Sonnet 3.5, but they’re looking correct. At the same time, I have no idea where this goes from here.
CJ Hess retweeted
The entire Amp team will be in NYC next week!
Come hang with the Amp and @tenex_labs teams on Thursday 10/8 to chat about building businesses in the post-agentic world.
No agenda, no speeches, just fellow travelers making IRL connections.
See you there! luma.com/2o2yus8t
CJ Hess retweeted
Amp team is in NYC next week! We'll be cohosting an event with @tenex_labs on Thursday! RSVP now to join!
The entire Amp team will be in NYC next week!
Come hang with the Amp and @tenex_labs teams on Thursday 10/8 to chat about building businesses in the post-agentic world.
No agenda, no speeches, just fellow travelers making IRL connections.
See you there! luma.com/2o2yus8t
It's orbing time (irl)
Come hang with @AmpCode and our team next week!
luma.com/2o2yus8t?tk=wPD0iE