“As I was floating on the surface, I realized something: When the turtle was swimming, it linked its movements to the movements of the water. When a wave was coming at him, he would float, and paddle just enough to hold his position. When the pull of the wave was from behind him though, he’d paddle faster, so that he was using the movement of the water to his advantage. The turtle never fought the waves. Instead, he used them.”
A notion from "The Cafe on the Edge Of the World" that I remembered (and found the actual quote thanks to a quick @ChatGPTapp deep research) - Cover from The Orb's "Auntie Aubrey's Excursions Beyond The Call Of Duty - The Orb Remix Project - Part 1"
I wish there was a simple way to temporarily switch "following-bubbles" on Twitter. I'd love to see what, say, sama or pg see when they open their feed. Does this exist?
The first print issue of Works In Progress is landing on people's doorsteps today!
Over the next week or so, our thousands of subscribers, everywhere from Alaska to Australia, should be receiving their first copy. We are trying to make the most beautiful physical magazine on the planet, and I am very excited about how we've started.
This issue has pieces on inflatable space stations (which you can read online today – t.co/kP1F46xUtO), the war between the Swiss and Japanese watch industries, the "great downzoning" in the world's cities, and on whether Ancient Greek and Roman statues were really painted as badly as museums claim (perhaps not). And much more.
I hope you all enjoy it! And if you'd like to get a copy of this edition, you can subscribe here and you'll be sent this most recent copy, along with the rest of your year's subscription as it comes out: t.co/hfTpdnnVjF
Seventy years after John von Neumann’s essay “Can We Survive Technology?”, the questions he posed feel more urgent than ever.
On Thursday 4 December, I'm chairing an expert panel discussion at the Hungarian Embassy in London on what that essay means to us now. You're invited! t.co/AW7kL60SaH
@TryAstroApp – Used Astro today to track, evaluate and adjust keywords for Pudding v1.0.1. Doubling down on those that work and replace those that don't. Very exciting :)
^ This is why I really don't engage in those FUD-dy "AI is just a bubble, it's too early to tell, it can't replace software engineers because blablabla..." conversations anymore.
AI works. Super, super well. It's one of the best co-workers I've ever had. What are y'all yapping about?
On the other hand I ALSO don't engage in the other, equally dark FUD-dy "AI works so well but it's soooo complicated to get it right. Oh my gawd, the only way you'll survive this is by taking my course" hidden agenda sales pitches.
AI works AND it is completely accessible to everyone.
There is NO secret sauce to get started. And there's no secret sauce and to get good either.
Second 🌕 moonwalk🚶🏻project successfully shipped to production:
-> Vibe code until prototype feels right
-> Turn prototype into markdown specification
-> Throw away vibe'd code
-> Keep specification
-> Feed specification to XP/TDD agents
-> Done
^ This is why I really don't engage in those FUD-dy "AI is just a bubble, it's too early to tell, it can't replace software engineers because blablabla..." conversations anymore.
AI works. Super, super well. It's one of the best co-workers I've ever had. What are y'all yapping about?
A good solution would not ignore reality but would try to figure out a way to keep the good and limit the bad - Josh proposes one such idea, at least for now: t.co/Qy8YdEf8TS
Hi Ethan, I published a Mac OS app recently that basically uses the editing history of a text as a uniquely human indicator.
I know that this is also not the full, final, solution, but it's a first experiment and can help figure this out.
0:00:00 – Explaining model jaggedness
0:09:39 - Emotions and value functions
0:18:49 – What are we scaling?
0:25:13 – Why humans generalize better than models
0:35:45 – Straight-shotting superintelligence
0:46:47 – SSI’s model will learn from deployment
0:55:07 – Alignment
1:18:13 – “We are squarely an age of research company”
1:29:23 – Self-play and multi-agent
1:32:42 – Research taste
Look up Dwarkesh Podcast on YouTube, Apple Podcasts, or Spotify. Enjoy!
We're proud to partner with @ENERGY and the Trump Administration on the Genesis Mission.
By combining DOE's unmatched scientific assets with our frontier AI capabilities, we'll support American energy dominance as well as advance and accelerate scientific productivity. t.co/8W2lgBRxoz
Anecdotally: Our AI productivity gains for production-grade code have increased by orders of magnitude once we got our autonomous XP/TDD agents working in @claudeai Claude Code.
Our sub-agents represent existing roles and responsibilities that we already had established and practiced. So writing them down required almost no effort.
Getting the main agent loop (CLAUDE(.)md) to orchestrate the agents took about a week to get right.
In total we have:
code-quality-reviewer
dependency-manager
locale-translation-specialist
test-coverage-analyzer
code-refactorer
frontend-design-specialist
qa-tester
db-migration-specialist
green-minimal-implementer
red-test-writer
Attached is a snapshot of how that looks in practice.
New Anthropic research: Estimating AI productivity gains from Claude conversations.
The Anthropic Economic Index tells us where Claude is used, and for which tasks. But it doesn’t tell us how useful Claude is. How much time does it save?
People making $300k a year from a SaaS do not need to make a few extra bucks selling you a “vibe coding masterclass”. This shit is tiresome. t.co/w3Dgo07xWa
Currently reading Atlas Shrugged and have enjoyed it very much so far. I'm half-way through, I'm on the edge of my seat but... I'm a bit worried about the whole the "industrialists and craftsmen disappearing and no one know where they went" part though. I am worried that John Galt is an actual person who is picking them up, building a kind of Shangri-La on some Billionaires island. :/
Just yesterday Alex and I were discussing the need for better day-to-day vocabulary even as a user of language models.
Thankfully, the people of Olmo 3 were way ahead and are putting a stake in the ground that open AI development should include access to all layers of a models' creation.
To truly advance open AI development and research, the entire model flow – not just its endpoint – should be accessible and customizable.
The model flow is the full lifecycle of an LM: every stage, checkpoint, dataset, and dependency required to create and modify it. By exposing this complete process, the goal is to engender greater trust and enable more effective adaptation, collaboration, and innovation.
Just yesterday Alex and I were discussing the need for better day-to-day vocabulary even as a user of language models.
Thankfully, the people of Olmo 3 were way ahead and are putting a stake in the ground that open AI development should include access to all layers of a models' creation.
To truly advance open AI development and research, the entire model flow – not just its endpoint – should be accessible and customizable.
The model flow is the full lifecycle of an LM: every stage, checkpoint, dataset, and dependency required to create and modify it. By exposing this complete process, the goal is to engender greater trust and enable more effective adaptation, collaboration, and innovation.
I'm noticing that once I've written one reply or post, I get into a momentum/frenzy and write another one and then another one.
It's like a sudden burst of ultra-satisfying energy.
I probably just really like hearing myself speak but it's also that I'm responding to what I hear YOU speak on here as well. So it's both: part egocentric echo-chamber of "my name is Alexis and the Universe revolves around me." and part conference of "wow, there's so many really interesting people with interesting topics on here."
I'm starting a collection of articles about workslop - will use this thread for continuous updates and links. t.co/5kghAOuUh6
AI ‘Workslop’ Is Killing Productivity and Making Workers Miserable
Also interesting that 404 Media is explicit about their human writing (emphasis mine)
404 Media is an independent website whose work is written, reported, and owned by human journalists. Our intended audience is real people, not AI scrapers, bots, or a search algorithm. Become a paid subscriber here for access to all of our articles ad-free and bonus content.
"What if progress in public science isn’t held back by bottlenecks in funding or talent – but by the institutions we run them through? Institutions are technologies. If we want to pursue new kinds of science, we need new institutional hardware." – Laura Ryan
Emphases mine, thank you @Ananyo for the heads-up!
New from @instituteGC: a proposal for Lovelace Disruptive Invention Labs
Is a better science possible, & how can we build it? 🏗️
Britain should pioneer a complementary model for research at the intersection of science & engineering – inverting core assumptions of modern public R&D
I believe the only way to “prove-prove” your question is by showing you how I arrived at my response to you here. The actual messy editing process and not just the final message.
For that, I’m working on an app that “proves your writing is human” called Pudding.
It’s for macOS and it’s an editor that also keeps track of my writing process.
If you’re interested, here it is on the app store:
This doesn't get old. It still mesmerizes me to see how well this works - even after months of running @Claudeai Code sub-agents in accordance with our XP/TDD practices.
And we haven't even touched Claude Code skills yet.
We definitely need to talk about responsible use of AI at work:
"...Of 1,150 U.S.-based full-time employees across industries, 40% report having received workslop in the last month. Employees who have encountered workslop estimate that an average of 15.4% of the content they receive at work qualifies. The phenomenon occurs mostly between peers (40%), but workslop is also sent to managers by direct reports (18%)."
Hi @screensdesign_ you seem like the person who would know this: Is it possible, for a macOS app on the app store, to also have videos instead of screenshots on the app's page?
I only saw "Upload your image assets in these formats..." in the App Store Connect UI (?)
Hi @screensdesign_ you seem like the person who would know this: Is it possible, for a macOS app on the app store, to also have videos instead of screenshots on the app's page?
I only saw "Upload your image assets in these formats..." in the App Store Connect UI (?)
Tired and zonked but happy that I, we, did this. Took us 10 years for cleaning up 81 years of family history and secrets and taboos and resentments. And blind spots. Never underestimate blind spots and the "operators" of the system that hide in them ;)
On the train heading home.
A heuristic for blind spots that's worked well for me:
"Trust your feelings, question your thinkings."
(Ahem, and yes, that's "thinkings", not "thinking" because I needed a better word. I'm also very proud of the whole thing)
Tired and zonked but happy that I, we, did this. Took us 10 years for cleaning up 81 years of family history and secrets and taboos and resentments. And blind spots. Never underestimate blind spots and the "operators" of the system that hide in them ;)
It’s the best model in the world for multimodal understanding, and our most powerful agentic + vibe coding model yet. Gemini 3 can bring any idea to life, quickly grasping context and intent so you can get what you need with less prompting.
Find Gemini 3 Pro rolling out today in the @Geminiapp and AI Mode in Search. For developers, build with it now in @GoogleAIStudio and Vertex AI.
The only reason it's such a small overall download count is that it's a total micro-niche app in the first place. But the ASO mechanics clearly work ☺️
To boost your app's keyword rankings, focus on downloads.
No matter if organic or paid.
These are the top "11" marketing channels to grow your app. t.co/h4y6HmRRX1
@matteo_spada I can confirm, even for a tiny app with only a handful of downloads, that within the last year, with just one round of @TryAstroApp-informed keyword revision, practically all downloads came from ASO. I was super busy at work, so I didn't follow up but, yes, clearly the way to go!
For me it was having to deal with work-slop internally, but I echo the sentiment. This is particularly annoying when it comes from people that are super talented because I'm like "duuuudeee...noooo"
Well, this is also exactly what I'm launching Pudding for. So at least I can lead by example and prove to the people that, for better or worse, I actually "wrote that myself".
I’m seeing candidates with incredible work backgrounds destroying their chances at getting hired because they sound like mindless robots.
Before you use AI in your job applications and interviews (which I’m largely against unless it’s edits or asked for), consider the following:
1) will other candidates use it too? And will my using it make me not stand out?
2) is the person reading it fairly advanced in AI? And will they know I’m lazily prompting AI for these answers?
3) does using AI in this process hurt me?
I test for AI skills in my interviews. I require AI usage (you wouldn’t be a culture fit on my team if you didn’t use it). And low-quality AI use is muddling the process and ultimately making it worse for many candidates, as recruiters will lean more and more on networks over oversubscribed AI-slop-filled application processes.
On my way to face THAT ONE biggest, deepest, ugliest family-related issue today.
You know that one at the heart of 99% of one's struggle.
Took me years to safely uncover it, build strength to deal with it, lots of pain, frustration and fear because it runs so deep.
So. Finally, I can't wait to go through with it. I don't fear it, I don't even feel particularly negative about it anymore. I feel more like it's obvious and totally to my advantage to get this over with.
This is getting silly…
Wanna buy the intellectual biography of one of the most consequential figures of the 20th century for less than $4!? t.co/DuzmtzMEK8t.co/x8tKbp8hlP
@Ananyo I wish there was also a signed collector's edition, hard-cover, letter-pressed - you know the fancy kind :) I would buy that for $250!
Considering buying a @MinimalCompany phone especially because of this blurb on their website. Would love to read this study and any others that investigate screen-time. Tips?
Currently reading Atlas Shrugged and have enjoyed it very much so far. I'm half-way through, I'm on the edge of my seat but... I'm a bit worried about the whole the "industrialists and craftsmen disappearing and no one know where they went" part though. I am worried that John Galt is an actual person who is picking them up, building a kind of Shangri-La on some Billionaires island. :/
I think it took me almost 20 YEARS to figure out Twitter.
I remember going to this podcast conference in 2006 at Hotel Pennsylvania. And one of the companies there was called Odeo. They were nice guys and I thought that was a really good name but podcasts were kinda whack.
A few weeks later they launched Twitter and I registered soon after. I made the mistake to pigeon-hole it along the lines of Dodgeball (probably because of the SMS part) though and didn't think much of it since. LOL.
On my way to face THAT ONE biggest, deepest, ugliest family-related issue today.
You know that one at the heart of 99% of one's struggle.
Took me years to safely uncover it, build strength to deal with it, lots of pain, frustration and fear because it runs so deep.
So. Finally, I can't wait to go through with it. I don't fear it, I don't even feel particularly negative about it anymore. I feel more like it's obvious and totally to my advantage to get this over with.
I was bored this weekend so I wrote a BBS for my PDP-11. Codespace is a maximum of 64K, and it's pure K&R v2. I'll let folks log in when it's more fully baked.
The hardest part was getting all of the menu selecting and highlighting and line drawing working using VT220 control codes. The rest is pretty vanilla :-)
You are offered a job with twice the current industry pay, but you must deal with older Unix operating systems and legacy applications all day long. Would you take it? #Unix #IT #Jobs t.co/EPH8Ofw99I
When you find yourself in a reading slump, consider revisiting the books that once captivated you.
Revisiting books that once sparked joy can reignite your love for reading.
Sometimes, stepping back onto comforting pages is the best way forward.
@readswithravi 100% agreed, that's exactly what worked for me - I restarted my reading with the two books that made me fall in love with reading as a kid: Enyd Blyton's "The Island of Adventure" and Jules Verne's "The Mysterious Island"
Okay, posting like this - without fear of what others may think - feels better than any note-taking tool I've ever used. I love Obsidian. And Notion. And even my own DIY apps. But this here, this Twitter-thingy. Wow. What a thrill.
I think it took me almost 20 YEARS to figure out Twitter.
I remember going to this podcast conference in 2006 at Hotel Pennsylvania. And one of the companies there was called Odeo. They were nice guys and I thought that was a really good name but podcasts were kinda whack.
A few weeks later they launched Twitter and I registered soon after. I made the mistake to pigeon-hole it along the lines of Dodgeball (probably because of the SMS part) though and didn't think much of it since. LOL.
Anyone remember the discussions we had when this landed in our orbit in... 2007? Over lunch 🍝 at Manganaro's up on 9th Ave?
And this was a few months before the iPhone came out.
For me, it was these four lines that resonated, and still resonate, with me:
there are notes in boxes that are empty
every room has an accessible history
you can search for sadness in new york
paths compete to offer themselves to you
While I take the liberty to ramble freely here, please always feel free to unfollow me at any time. No need to explain. I do I and you do you!
If my ramblings are just not what you need right now, PLEASE unfollow me immediately! Really. No hard feelings at all.
I've intentionally unfollowed (and then re-followed later) quiet a few authors here on Twitter in the past 6 months.
"Breaking up" with people, even for good reasons, can be hard or may even feel taboo.
I'm here to say to you: "No." is a complete sentence and is always, always, always an option.
I learned about this "law" when I participated in and co-organized BarCamps in the early 2000s.
"The Law of Two Feet" was, as far as I remember, the inoffical 10th rule of BarCamp:
All participants are responsible for their own experience and learning. If you find yourself in a session where you cannot contribute or learn anything, you may leave the room and switch to another session that is more interesting to you or initiate a spontaneous discussion outside of the sessions.
PS: If you haven't seen the other 9 rules, here's a quick reminder from t.co/KMi2bqBcJt
The Rules of Bar Camp
1st Rule: You do talk about Bar Camp.
2nd Rule: You do blog about Bar Camp.
3rd Rule: If you want to present, you must write your topic and name in a presentation slot.
4th Rule: Only three word intros.
5th Rule: As many presentations at a time as facilities allow for.
6th Rule: No pre-scheduled presentations, no tourists.
7th Rule: Presentations will go on as long as they have to or until they run into another presentation slot.
8th Rule: Presentations promoting specific comercial products or companies are discouraged.
9th Rule: If this is your first time at BarCamp, you HAVE to present. (Ok, you don't really HAVE to, but try to find someone to present with, or at least ask questions and be an interactive participant.)
Okay, posting like this - without fear of what others may think - feels better than any note-taking tool I've ever used. I love Obsidian. And Notion. And even my own DIY apps. But this here, this Twitter-thingy. Wow. What a thrill.
Yup, I'll say it loud and proud: I very much enjoyed reading NKS back in 2004. I want to read it again :) t.co/R1ObgemtsQ
I understand we have vocabulary, like NKS, that helps us understand and wonder that complex behavior can absolutely emerge from simple rules. For example Rule 11: t.co/jGocYRI8RH
Now, what about the opposite, converging, direction?
What do we call when things start messy and chaotic and, through say human intervention, end up ordered and genuine?
Yesterday, while preparing the AppStore submission for Pudding, I wanted to illustrate the gradual decrease of chaos towards a final, ordered result. I imagined what process Shakespeare may have gone through from first draft to the final "To be, or not to be, that is the question:"
And it's kind of a converging and messy process that starts with a lot of real-world noise ("Overheard on broadway: The cat is both alive or dead. WTF?") and then, by whittling and caring and putting in the time, gets cleaned and refined to its essence.
Does anyone know what vocabulary or research or models exist around this type of process?
What do we call the mess that we start with? What do we call the mess we make while trying to make it cleaner?
The problem with social media right now, summarized:
1) Rage bait works
2) Bots are plentiful
3) AI slop is everywhere
4) It doesn't feel super social anymore
5) Extremely addictive
6) Pressure to go viral
7) You compare yourself without noticing
8) Hard to tell what’s real
9) You forget why you opened the app
10) Everything is filtered
Yet, it’s still one of the greatest tools humanity has for building, connecting, and finding your people.
It's just far from perfect.
Hey Greg! What's been triggering me is definitely 2), 3) and 8).
Yet, against those odds, Twitter feels completely fresh and, you know, "real" again.
On a whim I decided to question what Twitter means to me two days ago.
I basically said to myself
"Okay, this place is kind of a shit-show. I can't change the bots, the grifters, the slop machines. So, what if I used it in another way? What if I use it to my pleasure and whims? What if, from now on, I treat it as my (somewhat) unfiltered notebook? What if I share what's on my mind without much editing or concern or fear of what others may think? What if, instead of down-regulating myself, I assume we're all adults here? If you don't like my stuff, please unfollow right away."
This is probably so obvious to you but I feel like I just discovered this whole new liberty and, dang, I haven't felt this alive in years! Honestly, it feels as electrifying as walking down a street in Manhattan on a early-June afternoon. Totally unexpected.
I'm loving it. After a day of posting practically anything that comes to my mind. Publicly. On Twitter. With almost no filter. Feels crazy liberating and energizing.
Is this what people do here? Is that what Twitter was all along??
"The Law of Two Feet"
While I take the liberty to ramble freely here, please always feel free to unfollow me at any time. No need to explain. I do I and you do you!
If my ramblings are just not what you need right now, PLEASE unfollow me immediately! Really. No hard feelings at all.
I've intentionally unfollowed (and then re-followed later) quiet a few authors here on Twitter in the past 6 months.
"Breaking up" with people, even for good reasons, can be hard or may even feel taboo.
I'm here to say to you: "No." is a complete sentence and is always, always, always an option.
I learned about this "law" when I participated in and co-organized BarCamps in the early 2000s.
"The Law of Two Feet" was, as far as I remember, the inoffical 10th rule of BarCamp:
All participants are responsible for their own experience and learning. If you find yourself in a session where you cannot contribute or learn anything, you may leave the room and switch to another session that is more interesting to you or initiate a spontaneous discussion outside of the sessions.
3rd Rule: If you want to present, you must write your topic and name in a presentation slot.
4th Rule: Only three word intros.
5th Rule: As many presentations at a time as facilities allow for.
6th Rule: No pre-scheduled presentations, no tourists.
7th Rule: Presentations will go on as long as they have to or until they run into another presentation slot.
8th Rule: Presentations promoting specific comercial products or companies are discouraged.
9th Rule: If this is your first time at BarCamp, you HAVE to present. (Ok, you don't really HAVE to, but try to find someone to present with, or at least ask questions and be an interactive participant.)
Actually that is because I'm planning to follow up on those notes with myself using the replies. I did that yesterday on some old tweets of mine and that felt surprisingly good and useful.
I'm loving it. After a day of posting practically anything that comes to my mind. Publicly. On Twitter. With almost no filter. Feels crazy liberating and energizing.
Is this what people do here? Is that what Twitter was all along??
This is all, of course, directly related to...
a.) me wanting a way to show my writing was, for better or worse, written by me and not AI
b.) the amount of AI work-slop that's started to land on my plate and for which I need some kind of filter soon
On the flip-side, I also have started to add "PS: Oh, and by the way, this is a hand-written comment!" when I post here. But that doesn't feel right (or believable) either.
My current experiment is a kind of "proof-of-sweat" editor where the messiness of writing is what makes it human. Here's a sneak-peek from earlier today:
Excited! I just submitted Pudding for App Store review.
It'll be a one-time payment but if you want to quickly beta-test this for free, DM me!
This is all, of course, directly related to...
a.) me wanting a way to show my writing was, for better or worse, written by me and not AI
b.) the amount of AI work-slop that's started to land on my plate and for which I need some kind of filter soon
To my total surprise and against all my prejudices, that Strengthfinder Gallup assessment I reluctantly took three years ago completely changed my personal trajectory and was some of the best spent money, if not the best money spent, in 2023.
I gotta thank my friend Tobi Eichenwald who had to poke me a few times to take the assessment.
I've also decided to use twitter as my personal, albeit public, notebook. So stay tuned for incoming, unfiltered stuff! :)
Actually that is because I'm planning to follow up on those notes with myself using the replies. I did that yesterday on some old tweets of mine and that felt surprisingly good and useful.
To my total surprise and against all my prejudices, that Strengthfinder Gallup assessment I reluctantly took three years ago completely changed my personal trajectory and was some of the best spent money, if not the best money spent, in 2023.
In a world that assumes everything’s written by AI, I made Pudding to let you prove the words are yours. It captures your writing as it actually happened — every edit, pause and rewrite — so you can replay the whole process and hand anyone undeniable proof of your work. Peace of mind for students, writers and freelancers who get asked, “did you really write this?”
@SpringStreetNYC@pickover Too kind sir! But much appreciated. In this world of social influencers and AI generated insights, please remember there’s human authors out there whose passion and graft gets this material out there (to be ripped off without credit by LLMs and YouTube grifters).
Absolutely – I, too, and too often, get turned off by the grifters.
Yet no one can steal or copy your talent, passion and time you put in. As a reader I would love to make this concrete: Here are three ways your writing has affected me personally:
1. The quote below from your book is one of my most cherished discoveries while reading. It shows how much you care and how much professional nuance you bring to the table. It made me pause, smile and reminded me of how magical the human condition can be. Honestly, I believe that Johnny is lucky to have you as his biographer.
2. "The man from the future" helped me understand and consequently appreciate the person and the genius of John Von Neumann. Equally though, it piqued my interest into you as an author. The only other author who has had this effect on me so far is Stewart Brand after reading his work.
3. Speaking of which and case in point: 10 minutes ago, I ordered the hardcover version of your book on top of the kindle version that I started with. I will not hesitate to order more copies as presents. And: Consider anything new you publish in the future already purchased.
Ramble, ramble. This is all to say: Love your work. Can't wait for what you do next. Johnny got lucky. Fuck the grifters.
PS: For better or worse, this post is 100% hand-written :)
@pickover Your regular reminder that an excellent intellectual biography of the man exists that focuses on what he actually accomplished for humanity through maths, rather than what others said about him t.co/lQiiHZFq6r
@Ananyo@pickover Ananyo's biography of John Von Neumann is so good that it landed in my top-5 all-time favorites. Totally unexpected. Again, thank you for it :D
Hey all! I wanted to share a kind of new development process I’ve been using lately — I call it "Moonwalk."
Basically it's:
Check out a spike branch (git checkout -b reconciliation/spike/mt940)
Vibe code fast (and with vim and vigour!) until it feels just right
Generate a spec based on the final version
Switch back to main (git checkout main)
Check out only the spec (git checkout reconciliation/spike/mt940 -- /job-stories) <- This is the moonwalk!
Let the "professional" agents and sub-agents write production-grade code based on the spec
It's so effective because I can leverage Claude Code for both: Quickly build a working vibe-heavy prototype spec AND write production-grade code with agentic/sub-agentic quality.
For the vibe code part, the standard Claude Code top-level agent is more than sufficient. It's great, no need for time-costly delegation to sub-agents here.
For the production-grade part, since we're an XP shop with a pretty solid TDD process, sub agents make total sense. We turned our existing process and skills into a bunch of sub-agents that work smoothly together:
We've got...
green-minimal-implementer(.)md
code-quality-reviewer(.)md
code-refactorer(.)md
db-migration-specialist(.)md
dependency-manager(.)md
frontend-design-specialist(.)md
locale-translation-specialist(.)md
red-test-writer(.)md
So, yeah, that's what I wanted to share with you. Try it out and let me know how I can help :)
Cheers!
PS: For better or worse, this post is 100% written by hand
PPS: My friend Alex Fedorov reminded me that this is pretty close to the spike/exploration workflow we used to do at Pivotal Labs. I think he's right.
cc'ing @claudeai and @Lovable because I feel like it's a nice connection between the two of you <3
@ChatGPTapp PPS: Here's another before/after illustration for how we may think about AI, AGI and it's impacts (Originally from Rem Kolhaas' Delirious New York)
Before 1930: What we think this is
After 1930: What the future actually is and means t.co/gMaXevHaWy
@ChatGPTapp Love the tagline:
"Buy a cozy cottage in our steel constructed choice lots, less than a mile above broadwy. Only ten minutes by elevator. All the comforts of the country with none of its disadvantages." – Celestial Real Estate Company
▶ Watch on Twitter
allenai.orgOlmo 3: Charting a path through the model flow to lead open-source AI | Ai2Our new flagship Olmo 3 model family empowers the open source community with not only state-of-the-art open models, but the entire model flow and full traceability back to training data.↗ allenai.org
404 MediaAI ‘Workslop’ Is Killing Productivity and Making Workers MiserableAI slop is taking over workplaces. Workers said that they thought of their colleagues who filed low-quality AI work as "less creative, capable, and reliable than they did before receiving the output."↗ 404media.co
betterup.comWorkslop: The Hidden Cost of AI-Generated Busywork | BetterUp LabsAI-generated ‘workslop’ is flooding workplaces. New research reveals how leaders can fight back and reclaim productivity.↗ betterup.com
Harvard Business ReviewAI-Generated “Workslop” Is Destroying ProductivityDespite a surge in generative AI use across workplaces, most companies are seeing little measurable ROI. One possible reason is because AI tools are being used to produce “workslop”—content that appears polished but lacks real substance, offloading cognitive labor onto coworkers. Research from BetterUp Labs and Stanford found that 41% of workers have encountered such AI-generated output, costing n…↗ hbr.org
X (formerly Twitter)Andrew Gallimore (@alieninsect) on XPsilocin slowly nuzzling its way into the 5HT2A receptor until it finds a comfortable position deep inside...
And therein the magic happens...
(molecular dynamics simulation from Bright Minds Biosciences)↗ x.com
en.wikipedia.orgRule 110 - Wikipedia↗ en.wikipedia.org
Chances With WolvesEPISODE 585 — Chances With Wolves↗ chanceswithwolves.com
xkcdOpen Source↗ xkcd.com
