Thriving on chaos with A.I. in a decentralized vs centralized 25 year dog fight wave of Future Shock and awe. 5th500 & Little Tech
United States
Joined April 2008
- Tweets2.2K
- Following609
- Followers594
- Likes411
Pinned Tweet
Forget the Browser Wars. Get ready for the Metaverse Wars.
Mark my words. The Metaverse Wars have already started. The battle for who gets to make the rules - to set the standard - is occurring right now.
And if you think that’s a trivial thing, you don’t know your history. The right to ‘set the standard’ was something Bill Gates was willing to nearly destroy Microsoft for.
He conjured up Internet Explorer as a brazen response to @pmarca at Netscape, and then used it to commit digital war.
In the process he got deposed by the US government, called ‘evasive and unresponsive’, and we got a glimpse of the true Bill Gates.
But, Microsoft survived and won the Browser Wars, and as a result, they ended up with a commanding position for the Web1 era - the benefits of which were felt across all lines of their empire. When you are the standard, you make the rules. Sure, other companies might also participate. But just as in the aristocracy, when a Duke was always ‘at home’ in any lesser nobleman’s house… the benefits of being the incumbent are real.
So with the stage set, let’s take a look at the emerging factions in the upcoming Metaverse Wars, their actions so far, their possible strategies, and the chances of their supremacy.
Let’s start with team Meta. Led by the brisket smoking, cage fighting Mark Zuckerberg, they were first to strike their colors, changing their company name to Meta, and spending billions on the Oculus Quest and Horizon Worlds.
Despite being first to the scene, with infinite money and a user base that includes literally everyone and their gran, so far their product appears to be a hot mess that no one really likes or uses. VR pioneer John Carmack left Meta citing slow progress and internal confusion.
Notable with Horizon Worlds, is the fact that you need to be logged into your Facebook account and using your Quest. It’s not browser based. Quite the walled garden - no doubt, it’s because Meta wants its 30% on all transactions done on its system. It's tithe, if you will.
While the general trajectory of Meta as a brand of late has been to get more byzantine, more filled with censorship and not better, only a fool would write them off.
They’re likely to strike an Unholy alliance with Unity, to make games for the Quest. They’ll be able to bring substantial entertainment force to bear, to create traction - plus the ability to target almost every person with a smartphone with their offers.
The Quest 3 drops later this year, which will offer blended ‘Mixed Reality’ experiences.But will that be enough to lure a critical mass to don expensive Oculus headsets and dwell in Zuck’s walled garden? We will have to wait and see.
Next, is Apple and their Alliance for Universal Scene Description. (My first thoughts - holy shit that’s a name as boring as the Bank of International Settlements, which means they’re not fucking around.)
And, just like the BIS, theirs has been a more guarded approach. At WWDC we had the launch of #SpatialComputing as a term- a strategic shot across Meta’s bows, presenting an alternative to the Metaverse concept.
Together with their high end Vision Pro headset (snorkel sold separately), Apple are presenting an initial offering aimed at those with the money to live futuristically.
What’s more significant, is who else is in the Alliance. Adobe? Autodesk?
If you’re a designer or creator reading this in your co-working space, sit up and take note.
Because my hunch is that before too long, you’re going to be recruited to join the Metaverse War on the side of the Alliance for Universal Scene Description (Glory be to USD! Hoo-ra)
Yes indeed. I imagine the USD will make a play for the creatives, the ones who grew up in Apple’s walled garden using Illustrator Photoshop and Autodesk, and enlist them to make the slickest Metaverse with glossy production values.
And honestly, that strategy might just work. If there is one walled garden that I still can’t quite quit, it’s Apple’s.
So expect to see Adobe updates that offer spatial computing functionality, as well as seamless cloud uploads to the USD compliant metaverse.
Will you be able to upload your work elsewhere? It remains to be seen, but my hunch is that it’ll be a lot less seamless, so just be a good creative and stick with your tribe.
We’ve explored the two emerging factions in the Metaverse War. We’ve conceded that both will have very credible offerings. But the one thing that neither faction will be able to truly offer?
An actual Metaverse.
A true metaverse is a single, universal virtual world. Interconnectivity and open-ness is key to the entire thing. I don’t think that either Apple or Meta will be able to share users. Each of them views us as users on its system. It’s their turf, where they get to impose their tithes and their TOS. No matter how slick the package is, they will never be able to offer us the real deal. We’ll have to see how many of the dev community choose to bend the knee and swear fealty to one of these Centra-Metaverses. We certainly won’t be.
Which leads us to the third faction in the Metaverse War, and likely the one that the Boardrooms in Silicon Valley are most afraid of.
Us.
The Decentralized Metaverse. Built by struggling artist devs, doing it for the love, not the money. This Metaverse is more likely to be made using true universal standards for practicality’s sake.
It will use distributed peer to peer frameworks like @holochain from @artbrock to ensure democratic principles and data integrity are baked in, and without the tithe to the overlord.
To get there, we might well need to do some jerry rigging - (it’s possible to build an affordable browser-based VR headset these days ).
We might need to do without the Zuckerbucks, and without the slick production values us Apple / Adobe folk have come to require.
It’ll be hard, but it is so important that we make the effort early on, to create a decentralized and open Metaverse.
It’s a chance to restart the internet project, in line with the beautiful VR cyberdream we’ve all shared since the late 80s. A chance to throw off the yoke of Big Tech, who want us to rent their Metaverse like share-croppers, and use technology to mould and monitor us in ever-more intrusive ways.
This is a call to arms, and if you’ve got this far in the article there’s a good chance you already know which side to choose.
mike mcclintock retweeted
LLMs will end up like Python did in software engineering: as a glue between different disparate systems. Eg when you want to connect a C++ app & a Java app & a web server you use python to wrap it all up. LLMs will similarly wrap different AI systems.
mike mcclintock retweeted
Everyone is slowly coming to this realization, and I assure you, no one is running multitudes of agents overnight. No one that is doing anything of substance at least.
There _are_ people pretending to be scientists, or fully caught up in their drug infused AI overdose, that think their slop machines are changing the world. They're not tho, and they're just wasting a bunch of money and compute to create a lot of LoC that will just get thrown away.
The state of the art is still "can we even one shot a production quality patch that we wont regret later", and its rarer than you'd expect based on discourse.
Talking to smarter folks than me, I'm convinced many of the AI folks in my timeline are full of shit.
Nobody is "running 20 agents over night" and building stuff for actual users. Maybe some are building internal tools or disposable software. Maybe.
But building software people like using? That doesn't get hacked on day one or blow up after the 3rd user? Nope.
I don't even understand what that's supposed to look like. Do you work out a 57 pages document that perfectly describes what you want to build and then summon 14 agents and have them run wild for 6 hours? And what comes out on the other end isn't a broken pile of shit?
Nope. Not buying it.
PS: it may also be that I have an IQ of 82 and can't figure it out.
mike mcclintock retweeted
The most powerful passage from Zen and The Art of Motorcycle Maintenance by Robert M. Pirsig:
mike mcclintock retweeted
Huge thanks to all the open source projects that've made a lot of the tech we rely on in the world possible:
Linux
Git
FFmpeg
PyTorch & TensorFlow
Apache & Nginx
MySQL, PostgreSQL, SQLite
Chromium & Firefox
GCC & LLVM
Docker & Kubernetes
Also, all the open-weight LLMs... and all the programming language interpreters, compilers, and frameworks.
I can list 100+ more "niche" open source projects I've used & loved (e.g., OpenCV ❤️). But I just wanted to give a continued big thank you to the open source community for everything you do for humanity 🙏
mike mcclintock retweeted
Phoenix LiveView (1.1) is getting colocated JS support!
While JS commands can cover many scenarios, sometimes you just need a sprinkle of JS in a certain component.
I'm excited for this to come out soon!
#MyElixirStatus
ALT An example of ColocatedJS that registers a custom `copy-to-clipboard` event in a Phoenix LiveView function component.
ALT An example of the new ColocatedHook in LiveView, in which an input has custom JS that applies a specific phone number format to the value.
Not all crypto moms wear capes
I'm thinking about a potential exemption from registration that would allow firms to use DLT to issue, trade, and settle securities. A sketch of the idea is here: sec.gov/newsroom/speeches-st…
This is way —
A Bootstrapper's Guide to Re-Industrializing America
I’m a big fan of focusing on things that are within my control. This article is about real actions we can take while we’re waiting for lawmakers to get their shit together.
@andrewmccalip, @zanehengsperger, and others have been writing great articles about reindustrializing, and I wanted to say HELL YES to everything I’ve read so far. I fully believe that their policy ideas need to be implemented in parallel, but I’m not one to wait around for lawmakers to make my life better.
This article isn’t about starting new billion-dollar foundries, refineries, or nuclear power…that is way over my head. I’m talking about what a couple of guys can do, in their own town, and support a family (and a healthy number of employees and their families). Rely on yourself. Just go do stuff.
The CNC manufacturing boom in the 90s created thousands of entrepreneurs…
**Example 1: CNC disrupts traditional talent**
The 1990’s were an awesome time for US manufacturing. I was young, but I was able to experience the “modern” CNC wave first hand. My first hourly job was at a small sign shop in Minden, Nevada. The owner had built the business on his artistic talents: He could hand-paint about 20 different “alphabets” (fonts) by hand, from memory, on any surface from glass to the door of a pickup truck. Yes, it took a 6-pack to get the lines nice and straight, but damn he was good.
Mac was a helluva artist, but he also knew that in order to keep up with demand, he needed to modernize. He invested some money in an early vinyl cutter/plotter so that younger staff (like myself) could make some of the simpler signs and graphics. The Windows 3.1 CNC interface was primitive, but suddenly I could create an end product that didn’t require a lifetime of practice. The side effect of this technology was that it made a previously talent-dependent industry available to anyone with a PC and a checkbook.
Sign shops popped up out of nowhere. Our tiny town had 3 or 4 within a couple years, and there were more entrepreneurs running vinyl plotters in their garage competing with us. Our owner’s 20 year head start evaporated within a few years. However, operational excellence, a good brand, and outstanding service protected him in the long run.
The point I’m trying to make is that new technology created a wave of entrepreneurs that was overall net positive to the industry and consumers. A bunch of people left their corporate jobs to start their own businesses, and that’s something I think we can replicate today. Something similar happened in the world of fabrication and manufacturing:
**Example 2: CAD/CAM makes manual equipment obsolete**
In the late 90’s CAD/CAM systems were developed that actually worked and didn’t require manual operations, navigating punch card programmers, AS400 systems or even DOS. Suddenly fabricators saw that instead of hiring machinists with 20 years of experience, they could hire young guys who could navigate a Windows environment and spit out GCODE that actually worked on the new generation of machines.
Again, a wave of entrepreneurial types took a chance on this new style of fabrication and spun up businesses left and right. I’ve heard some ridiculous-but-true stories of equipment salesmen selling equipment from the back of limos, shop owners wrecking new Porsches because they were too coked up, and many, many divorces caused by “new money.” Once again, the catalyst for all of this was new technology that enabled new entrants to get into the industry. Fabrication and manufacturing was no longer something that was guarded with the prerequisite of generational knowledge.
…But then the CFOs/MBAs got ahold of things and threw it all away
The US saw an explosion of growth in machine shops, fabricators, and manufacturers enabled by new tech. It created a lot of growth and wealth, and many of these companies scaled quickly. The only issue is that as they scaled, MBAs and CFOs started to get creative and look for even more margin. Offshore labor could add a few basis points of margin, so why not send some work overseas? Seems innocent, ya? The rest is history. (Before you comment, yes, I know that offshoring was happening before the 90s, but I was busy playing Duck Hunt so I’m not going to get into it)
We can re-boom
I think it’s time for a 90’s-style equipment-led re-boom.
Machine shop owners that prospered in the 90s are retiring. They don’t want to invest in new equipment to be competitive, they are tired. They’ve been doing this shit for 30+ years…it’s time for someone else to take their place. I don’t see machine shop retirement or closing as a bad thing. I think it opens up some room for new entrants, using new technology. The current generation of manufacturing equipment is light years ahead of anything from 1995…or even 2015. (Caution, wild generalizations ahead, bear with me) The Human/Machine Interfaces are easy to use, CAD/CAM software is the best it’s ever been, equipment uptime is up and maintenance needs are down. There has never been a better time to invest in new equipment and go win some work.
I believe that if more brave entrepreneurs leverage this new technology and take the plunge into manufacturing, thousands more will follow. The reason manufacturing is hard is because many of the existing companies are saddled with old tech that is hard to learn, maintain, support and expand. A fresh start is what we need.
“But Jim, all that new equipment is expensive! How do I buy it?”
Use Other People’s Money - Specifically, The equipment manufacturer’s money
There’s a lot of talk about how the government should offer low-interest loans to manufacturers for reindustrialization, and I agree wholeheartedly. But who knows how long that might take. I did an SBA loan years ago and just the application process took 6 months. Who knows how long it’ll take for new legislation to make its way down to the guy trying to get the loan.
Something you can do now is choose a manufacturer that offers in-house financing (or has a finance subsidiary), like Haas, Amada America, or dozens of others. We’ve been able secure very attractive terms from in-house lenders, mostly because they exist to sell more equipment, they aren’t a bank. I’ve seen 0% down, 0% interest, delayed 1st payments, cash back…everything is on the table if you ask. Especially if that ask is at the end of the Sales guy’s quarter.
Manufacturer financing can be such a massive advantage for a startup, and it doesn't dilute your equity like a VC would (VCs shouldn’t be in manufacturing, too much pressure to grow too fast). You’ll need all the advantages you can get to compete against China.
“But Jim, how can a small manufacturer like me compete against China?”
Do what China does, but better
China will continue to be the lowest cost provider for commodity goods until we can outpace them with scale. That will take a while. Their cost of living is cheap, labor is cheap, their raw materials are cheap (subsidized?), and somehow they can ship something from China to Wisconsin for less than I can ship something to the next state…wtf (we need to eliminate the Terminal Dues System, but that’s another topic).
So, I suggest we don’t compete with them on mass-produced commodities, yet. That will take loads of scale that is beyond the scope of this article. Instead, we can all pull a few pages from their playbook and use them to win work here in the US.
**Follow up.**
China is pretty damn good at being responsive. If you send a quote over, any time of day or night, you’ll usually have a response within an hour, usually minutes. Go try aliexpress if you don’t believe me, you’ll get about 30 emails per hour until you buy. And once you pull the trigger, they will be cutting chips before some guy in Ohio answers the first email.
American manufacturers have to focus on being responsive and having the best communication in the world. I don’t care how you do it, just respond.
Recently, I wanted to get a quote on a yard cleanup at my house to prep everything for winter. I called/emailed six companies that I found on Google. Four never responded. One scheduled a time nine days later to come give me a quote and didn’t show. One looked on Google maps to see how big my yard is, called to ask a few questions about the property, and gave me a quote during the same call. They won the job. The work is out there if you call people back.
**Your customers work 24/7, so should you.**
Respond via text, email or DM as many hours of the day as you can. If you need to sleep, hire someone, use an answering service or virtual assistant service. They pay for themselves on the first job. Modern customers don’t work 9-5, Monday-Friday. We get orders around the clock, 7 days a week, and I know the demand is there for other manufacturers too. Personally, I do the majority of my shopping for new equipment on the weekends when I have some time to think and browse websites. It’s unfortunate that I typically don’t hear back from anyone until Tuesday or Wednesday, usually later (if at all).
**Put your price on the website!**
CNC guys: Lean into instant quoting software, it’s out there and it’s getting better all the time. We write our own pricing software, but there are some off the shelf options that will get you started.
Product manufacturers: Yes, I know that there is a huge percentage of manufacturing that is custom and needs to be quoted, but if you can, put your prices on the website (Haas does a great job of this). If it’s fully custom, put a price range or a variety of examples. It’s a win for your customers because they can get their expectations and budgets aligned with what you offer, and it’s a win for you because you aren’t wasting time quoting products for people who can’t afford it.
**Go faster damnit!**
8 week lead time ain’t gonna cut it. Run 3 shifts, run lights out, whatever you gotta do. The work is out there if you are fast, I promise. Today, if you are 8 weeks out, China will be 2 weeks or less (shipped). You gotta do it in 5 days. The nice thing about speed is that it insulates you from price. If China is going to take 2 weeks and it costs $1000, you can do it in 5 days and charge $2000 and still have loads of demand. 80% of the buying population is impatient, bad at planning ahead, or both (myself included).
“Do you ever think about the fact that Dominos does customized built to order products with a lead time of 15 minutes, all through an instant pricing configurator based website? I do.” - @andrewmccalip
**Document your process!**
Make it repeatable. It’s sooo much easier to train this way. Training manuals, checklists, operations manuals, written policies. It’s a big investment of time at the beginning, but it’ll make your entire operation run smoothly for decades. You need a smooth operation so that you can be responding to customers and making things faster.
**Compete on service.**
Personally, I spend more to have good service (especially when I buy equipment), and there are many others out there like me. Sure, the product might be 15% more expensive, but if it comes with 200% better service, that’s the product I’m buying. As manufacturers, we need to be long term partners with our customers, and that means focusing more on service before and after the sale. This is an area where China can’t compete because once the sale is complete, they are moving onto the next target.
American manufacturers need to take a few notes from the hospitality industry: Make it right for your customers when something goes wrong, overnight spare parts when they need them, beat deadlines, follow up a few weeks later and see how things are going. I’d love to see American manufacturing positioned as “we aren’t the cheapest, but we’re the best.” Yes, we need to find efficiencies to reduce our prices where possible, but we have competitive advantages in front of us in the form of speed, service, quality, support, and communication.
The reason China is winning work is because they are cheap, decent quality, and pretty fast. If we can’t compete with cheap, we sure as hell can win on everything else.
But I can’t get the material for what they are selling finished products for!
I know, no one can get material for what they are getting it for. It’s pretty wild how cheap things are over there. But, that’s the reality so stop complaining and find a way to move forward. In addition to offering the best speed, service, quality, support, and communication, we can do a few things to be smarter about our operations.
**Co-op buying systems (regional).**
Manufacturing is incredibly fragmented in the US, and I think we can band together a bit better and streamline how we source raw materials. What would it look like to join together with 10 other machine shops in your city and purchase billet using a co-op entity? You’d have 10x the buying power, and potentially lower freight costs if you can order in larger quantities to a central location. And hopefully you can make a few friends in the industry and support each other. I know it works for us.
Also, there are other raw material vendors besides McMaster. Don’t let their pricing dissuade you from entering the market. They are the best company on the planet for getting you what you need, delivered the next day, but when it comes to buying commodities, don’t rely on their pricing when running your margin forecasts.
**Build your business where the raw material is.**
This is what China is really good at. Want to start a sunglasses factory in China? You better do it in Wenzhou. There are over 2,000 eyeglasses companies in that region. You’d be starting with an entire supply chain ready to go, including the workforce.
In the US, if you need lots of wood in your product, you might want to be in the Pacific Northwest. Need lithium because you’re making batteries? Go to Nevada. Steel? Try the Midwest. Need lots of power? Move near a dam or solar farm or nuclear plant. If you can’t compete because you decided to start an organic Aloe Vera lotion company in Michigan, I don’t feel bad for you. Try Arizona.
**Make things that are heavy, complex, custom.**
The more size, complexity, and customizability your product has, the more insulated you are against China. Don’t focus on making products that are tiny, simple and easy to duplicate. They will spin up a copycat factory and wipe you out.
Instead, make big things that are challenging to ship halfway across the world. DHL’s maximum shipping weight for Economy Select is 154lbs. The fast version (Express International) is 66lbs.
**Make things that are complex and need a lot of engineering.**
Make things based on customer specifications, things that need collaboration between the customer and the manufacturer. That’s where US companies can’t be beat.
Beyond bootstrapping - where I hope the government can help
I know I said at the beginning that I wasn’t going to focus on long-term changes related to government policy, but there are a few things that I’d like to mention to help guide lobbyists and lawmakers:
**Win over lawmakers with job creation.**
Lawmakers love jobs. As manufacturers, we need to talk more about creating the next wave of high tech, high paying manufacturing jobs, and talk less about automating everything towards lights out. Yes, we’ll have robots and automation, but we’ll also have lots of great people around to engineer them, program them, and work side by side with them. The story has got to be about people, not robots. You are 100x more likely to get a grant or tax abatement by talking about job creation vs. a dark warehouse full of conveyors.
**Build “Cash for Clunkers” v2.0 - but for equipment.**
This is a big one for me, and I bet a few of you are connected enough to help me pull it off. Let’s do a government-backed cash for clunkers program, but for outdated manufacturing equipment. If you really want to reindustrialize the US, we gotta get rid of CNC machines still running on 5 ¼” floppy drives (or punch cards). I’d love to see some sort of credit towards a new machine if you “push pull or drag!” in your old machine, or show proof of recycling. Hell, we probably don’t even need the government to help here, a smart equipment manufacturer can just steal this idea and run with it. Amada, Haas, I’ll be first in line if you wanna test this out.
**Government-backed loans to equipment manufacturers. **
Giving direct loans to startups is risky, and the government doesn’t have the bandwidth to vet all applications properly. Remember the PPP program during covid? That didn’t work out so good, except for the fraudsters who got Lambos. Instead, what if we gave equipment manufacturers (Haas, Fadal, Mazak, Amada, PEKO, Parker Hannifin) government-backed loans that they could use to offer better loan terms to the end user?
Their sales teams are great at vetting customers, and they aren’t going to be as vulnerable to fraud as the government would. Every equipment sales person I know physically visits the operation, learns the business, gets to know the operator. The idea is to use the equipment manufacturers as a way to distribute the funds, in the form of cheap loans.
Conclusion
If you sell good equipment with good terms and flexible financing, people will build a business around that equipment. I have seen it happen in 1995 (Modern CNC equipment), 2005 (iPhone Apps), 2015 (CNC plasma tables) and now in 2025 (next-gen manufacturing equipment).
Haas helped create more manufacturing entrepreneurs in the last 20 years than any other company I can think of, and I believe we can scale that model 100x.
If we can show young entrepreneurs that manufacturing is actually really fun (Factorio in real life), and it doesn’t require 20 years of experience (due to modern equipment and software) we’ll reindustrialize from the ground up, 2,500 square feet at a time.
If you can pronounce any of the names in your Lions mock draft, redo it. #onepride that is Brad Holmes secret.
mike mcclintock retweeted
It is clear to me at this point that the true torch-bearer of the IDW is Bret Weinstein. While Shapiro, Rubin, Harris, Kisin, and Murray have all turned full heel, he continues to forge ahead. Adjust your trust and respect levels accordingly. Bret is for real.
Vibe coding by non technical founders might be the best productivity hack since the invention of the assembly line. By creating something tangible to iterate through at lightning speed, the non technical founder creates a feedback loop that cannot be duplicated by trying to communicate a concept to a developer.
Communication between product owners (the idea guy) and developers is the big bottle neck and has been forever.
Communication between a non coding founder with an LLM and back to his or herself can produce a years worth of refinement of the idea in the span of 50 hours.
Then have a developer use AI to quickly build the product with clarity.
AI is the best communication tool between the technical and non technical ever invented. #buildinpublic
mike mcclintock retweeted
TBE is a foundational framework for looking at the Universe through geometry.
As a freebie result from TBE - here is possibly the first computed image of a "quark pair" and their gluon flux tubes.
A closed loop with a twist in 4space appears as a confined knot with a partial twist in 3space
Twists can be interpreted as charge - unresolved strain
When you project a twist in 4space to 3space you observe the shadow of the twist - the fractional charge
Quark bags - Baryons/Hadrons - are closed strain loops in 4 space with twists.
This is a Pion - it a closed loop in 3space - but it meets at a discontinuity. The purple "dot" represents a point where the strain tensors are balanced in 4 space but asymmetric when projected into 3 space ... this happens where the ribbon has undergone 1/3rd twist - the remaining part of the ribbon has a 2/3 twist.
The Pion exists at the purple dot - and the ribbon translates into the gluon flux tubes of QCD
Mesons and quarks under the belt (some good candidates for Charm and Top) ... Going to tackle Baryons tomorrow.
Further results are coming ...
@ericweinstein @TAndersen_nSCIr @johncarlosbaez
Who are these shills propagating the erroneous math that a tariff will be exponentially marked up at every stage of the value chain? Even if you accept the claim that each stage is marked up 100%, the tariff in the worst case is passed along without a markup.
The example seems to be an iPhone where a company pays $100 to the factory, then sells to the wholesaler at $200. The wholesaler sells to the retailer for $400 and the retailer sells to the consumer for $800.