I ran across this video a few days ago and couldn’t stop watching it.
It’s about something ordinary & boring, a plastic gas lighter. But it changes how one thinks about manufacturing.
That lighter in so many of our homes, holds pressurised gas. It has over 30 microscopic parts, has to pass international safety codes, & travel 10,000 miles by sea, & the total cost of doing all that, materials, labour, freight, every middleman along the way, comes to fifteen U.S cents.
So how does anyone make money on this?
Turns out almost the entire world’s supply comes from one place: a county called Shaodong, in China’s Hunan province.
It wasn’t always there.
But today, Shaodong has 114 lighter-related companies packed into the place & between them they source more than 200 different components from each other, all within a 20-kilometre radius. They supply something like seventy percent of the world’s disposable lighters. And the industry alone employs over 80,000 people locally.
Nobody there is winning on cheap labour anymore. They’re winning by shaving a thousandth of a cent off the thickness of a plastic wall, or redesigning a base so a few thousand more units fit into the same shipping container.
It took my thoughts back to an old professor of mine, Michael Porter.
His 1980 book, Competitive Strategy, is still the 1st book most MBAs read, the one that gave the world the Five Forces and basically invented modern strategic thinking.
But there’s a quieter piece of his work, on industrial clusters, that never got nearly the same attention, and it is the one that explains exactly what is happening in Shaodong.
His argument was that nations and regions rarely win because of cheap inputs. They win when rival firms and specialist suppliers crowd into the same small geography for long enough that they keep pushing each other past what any one of them could manage alone. He found it in the Swiss watchmaking towns of the Jura, in the German printing press industry and in Italy’s ceramic tile and footwear districts (interestingly, it’s the SAME blueprint which built Morbi, in Gujarat, into the world’s second-largest ceramic cluster, now outproducing Italy by volume. I have posted before, about Morbi)
None of these started out as giants. The neighbourhood made them giants.
Which is exactly why it’s so relevant to India’s climb up the global manufacturing table
I’ve also attached a slide with this post that I saw recently and which shows us breaking into the top 5 manufacturing globally. (A quick reference check told me that we may not have overtaken Korea yet, but the trajectory’s clear)
That climb has happened on the back of scale: bigger plants, bigger parks, more FDI.
I should declare an interest here, because the Mahindra Group set up 2 of India’s first integrated, plug-and-play business cities, in Chennai in 2002 & Jaipur in 2006.
Both have been extremely successful. Chennai’s business zone alone today employs 45,000 people..
But I admit that we need to think differently.
A park brings in investors and hands them a ready plot, power, water & roads
A cluster is a completely different animal: hundreds of small, specialised suppliers, each obsessed with doing a tiny thing better than anyone else, feeding off each other’s presence for years until no outsider can compete with the whole.
I think that’s the work ahead of us now.
Not just more factories, and not just more parks.
Policymakers & developers like us need to start consciously pulling as many of the inputs and resources a sector needs, the toolmakers, the component suppliers, the testing labs, the logistics specialists, into the same neighbourhood.
Shaodong and Morbi both got there by accident, one town stumbling onto a way to shave a thousandth of a cent off a lighter wall, the other discovering it had the clay and, later, the gas pipeline for tiles.
We don’t have the luxury of waiting for accidents anymore.
We need to do it on purpose
My drug dealer is a hypocrite! @DarioAmodei#Anthropic product is COCAINE.
1. weeks ago, with no compute to serve it they campaiged for regulation on Mythos.
2. Then @elonmusk ='collosal' compute and they want the model out.
3. but wait, wishes came true - REGULATION.
You may hate the name but this is one of the best designed social programs in the world to actually address income gap over time. Pls use it if you can and are eligible.
Make sure every child under 18 & their families that u know download their @TrumpAccounts app today! All kids under 2 get $1000! Most kids 2-10 at least $250! And kids 10-17 a free lifetime 401k like acct that employers & others will add $$ to! Every child a shareholder! 🇺🇸🚀
For non tech folks
This is like Sachin coming out of retirement in the form of his life and Joining RR and RCB together.
Or Ronaldo Joining Manchester united.
Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.
@karpathy popularised the idea of a "second brain" — one system that holds everything a company knows, learns from every interaction, and gets smarter over time. Most companies are still debating whether to start. We're running ours in production. Here's how 🧵
I have a theory that if you were a teenager between 1998 and 2004, you accidentally became insanely high-agency. There were no streaming services. No $9.99 Spotify or Disney+ to get what you want. If you wanted entertainment, you had to become a cyber criminal.
Do you know how complicated it was to just get access to music, movies and porn? Every day after school I was bootlegging Dane Cook comedy sets, learning bodybuilding from obscure internet forums, and if my mom picked up the home phone line, the entire internet would disconnect. Shoutout to my fellow late 1980s kids, we unknowingly did navy-seal style training on how to hack the internet to get what you want.
Judging by my tl there is a growing gap in understanding of AI capability.
The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code.
But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along.
So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions.
TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.
Voice AI for Bharat is a completely unique problem statement, trust me.
We run 18,000 debt collection calls a day across India.
Hindi flipping to English mid-sentence. Speakerphone on a busy road. Auto-rickshaws honking. PSTN lines crackling at 8kHz.
Every call must comply with the RBI's Code of Conduct. Miss a FATAL violation — abusive language, sharing debt details with a family member, skipping identity verification — and it's a regulatory risk.
The industry audits 2–5% of calls. The rest go unheard.
We built an AI system to audit all of them.
19-parameter quality rubric. 7 FATAL flags. Auto-scored within hours of every call.
The architecture is ready. The compliance engine is live.
But the most critical question is still open:
Which STT + LLM combination can actually understand real Indian phone calls?
So we're running a hackathon.
50 real production calls. 5 vendor combos (Soniox, Sarvam, Deepgram, OpenAI + GPT-4o-mini or Claude 3.5 Sonnet). Blind evaluation.
You don't need to be an engineer. Even testing transcription alone is a complete submission.
Winner gets shipped to production at scale.
https://t.co/o2uTsoG8ws
@GodboleSandeep@dhavalkulkarni Because back in the day someone told him he's better off having a south Indian name in Bollywood vs a marathi name. Turned out fine! He was more popular as a music director than a singer!
@MarkJCarney@CanadianPM
I don't fully agree with your policies but thank you for #BC12. I feel respected as a legal immigrant and naturalized citizen of Canada.
Let's enforce it 100% and cleanup this mess. #MakeCanadaGreatAgain Thank you.
Everything is mined or grown
Everything normal 20 years ago is DEAD.
True democracy is democratising knowledge & technology at near-zero cost.
The hardest to fall will be institutions, tradition & central authority.
Massive friction is coming from the “surplus elite” — the socialist frauds who preached dignity and equality for all while hiding their racism, classism & NIMBYism behind a cultured veil of hypocrisy.
Power returns to the builders and growers — taken from the fake institutional stamps of credentialed elites.
@TigerWoods what are you doing sir?
YOU ARE MY HERO and you mean a lot to so many of us. Get it together for us if not for yourself & your family. Take care - you did the crime now do the time & lets get another 6 majors in the bag!
The stupidity of people!
This video is fake. Iran has hit a F35 US has confirmed an emergency landing.
So manythings are wrong here - people are still sharing like crazy
1. Shows a casual straight & Level flight. Even after multiple threats in proximity, any pilot would manuver in aggressive evasion, Break hard, high-G turn, barrel roll, or split-S to defeat the threat vector. They would
Dump chaff/flares aggressively while pulling max Gs to break lock. Minimum they could climb rapidly to increase distance. This guy is flying a video game!
2. The FakeThermal/IR signature problems- the entire airframe glowing uniformly hot, almost matching the engine exhaust plume brightness. Real fighters thermal profiles are very distinct: engines/IR signature hot, but fuselage and wings stay much cooler due to stealth coatings and design.
3. Bright central explosions occur, smoke trails, yet the jet continues flying normally with no wing shredding, control surface failure, nothing. A direct SAM hit on center-mass would almost certainly cause catastrophic structural failure — not a gentle trajectory.
4. Countermeasures — Flares are deployed, but in a way that doesn't align with real procedure - no timed bursts to spoof incoming IR-guided missiles, no immediate hard turn away from the threat - looks hollywood.