Andrew Opdyke retweeted
The highest quality video of the moon was just released… this is so beautiful.
Andrew Opdyke retweeted
After the Space Shuttle Challenger exploded in January 1986, killing its seven crew members, President Reagan appointed a commission to investigate.
Richard Feynman, already battling cancer and reluctant to join, accepted because a former student asked.
He quickly grew frustrated with the slow, formal hearings and NASA’s optimistic safety claims (1 in 100,000 chance of failure).
Instead, he talked directly to engineers, who revealed far higher risks.
The night before a key televised hearing, Feynman bought a C-clamp from a hardware store.
During the session, he took a sample of the rubber O-ring material from the solid rocket boosters, clamped it, and dropped it into a glass of ice water (mimicking the cold launch temperature that day).
After a moment, he removed it and showed how the rubber had lost its elasticity, it no longer sprang back.
He explained simply: at low temperatures, the O-rings couldn’t seal properly, allowing hot gas to leak and cause the disaster.
His live demonstration cut through layers of management denial and became one of the most iconic moments in engineering accountability.
In his personal appendix to the report, he famously wrote:
“For a successful technology, reality must take precedence over public relations, for nature cannot be fooled.”
Andrew Opdyke retweeted
Humanoids = stacked actuators. Pure and simple.
If you don't understand actuators, you don't understand robots.
Actuators are not solved: too heavy, too hot, too inefficient.
This is a mega thread on actuator design for humanoid robots, directly from the mind of @GoingBallistic5
Bookmark it for later 👇
Andrew Opdyke retweeted
I get asked all the time why we buy US-made (or Japanese) equipment vs. cheaper chinese machines.
"Dude, you could buy 5 of the chinese machines for the price of one of of the US ones!"
True, but it wouldn't necessarily mean more uptime, efficiency or profit. I buy US-made machines when possible because I need the reliability, service, support, and communication. It's more money upfront for better quality of life down the road.
My business model isn't compatible with chinese machines because:
- We are quick-turn. We can't afford hours of downtime let alone days or weeks
- We are square footage constrained. If I only have 600 square feet to place a machine, it needs to be the best, most productive machine possible
- We have high demand and run 24/7. The machines need to be spec'd to 120% duty cycle in a harsh environment
Models that can work well with chinese machines:
- Long lead times / patient customers. If a machine goes down for a few weeks, and it won't ruin you, go for it.
- Unlimited square footage. That way you can have "spare" machines in case one takes a poop
- Just testing out the model to make sure it works before making larger investments ( I see this a lot and I agree. Kinda like buying a harbor freight tool to see if you like it. When it breaks, go buy snap-on or whatever)
Buying high-quality US/Japanese machines has worked for me for decades now. In the long run, I save/make money with them because:
- The warranty is usually great, or at least I can buy a service program. With high-dollar equipment, it's nice to be able to get a factory tech on-site within a day to fix things vs. wechat messages or forums
- Good machines are full of good parts. Name brand motors, solenoids, controllers, bearings...stuff you can buy on McMaster or Grainger if needed.
- Factory spares. We keep critical spares on-hand, but if we need something weird, we can usually get it overnighted from the US factory or distribution center. Much harder with chinese stuff
- US/Japanese (and German too, I forgot about them) are usually way overbuilt and can handle a decade or more of abuse. The chinese machines are cheap because they cut corners, use thinner/lighter materials, cheap motors, and don't last as long. For some of our CNC machining, we use Matsuura MAMs. I know of a few in the wild that are 19+ years old and still churning out parts with tolerances measured in tenths-of-thousandths of an inch.
- Resale when it's time to upgrade: A good machine from a well known name will hold value much better than a chinese machine. Often times chinese machines aren't worth the rigging/moving costs once it's time to replace them
- Financing: It's 1000x easier to get a loan against a good US/Japanese manufacturer than a chinese machine. Banks are using the machine as collateral, so they want to make sure it's going to be worth something if they have to come get it. Equipment lenders know the market for brand name, popular equipment, so they are more likely to lend against it.
Today we use:
- Amada America lasers and press brakes - Brea, CA and Highpoint, NC
- Amada Japan lasers - Isehara City, Japan
- United Finishing - Statesville, NC
- Brother CNC machines - Kariya City, Japan
- Matsuura CNC machines - Fukui City, Japan
- Powder coating equipment - various companies in MN and WI
- Toyota forklifts - Columbus, IN
- Omax waterjets - Kent, WA
- etc
If you can make the initial investment pencil out, buying the highest-quality machine available ends up saving you money (and making you money) down the line. Little pain up front for a lot of gain down the road.
Andrew Opdyke retweeted
This would require deflation, which may be more damaging than inflation - it’s why resetting the price of gold lower in history always led to a severe economic contraction. The target should be zero inflation from now on. And real rates are too high.
Andrew Opdyke retweeted
The U.S. Department of Education reported that 42.7 million borrowers owed a combined $1.6 trillion in student loan debt as of 4/21/25. Of that total, nearly five million borrowers were in default (had not made a monthly payment in at least 360 days), while four million borrowers were in late-stage delinquency (91-180 days delinquent). Collections of defaulted student loans will resume in May of this year, ending the pandemic era pause that began in March 2020. First Trust: Factoid of the Day
Andrew Opdyke retweeted
First Quarter Bloomberg Economic Forecasting rankings are out. @BobStein_FT is a rock-star.
Andrew Opdyke retweeted
Senate Dem incumbents retiring in at least three purple states — Michigan, Minnesota, and New Hampshire — makes it very hard for Dems to take back Senate in 2026.
Andrew Opdyke retweeted
When @NickTimiraos speaks, markets need to listen…
The debate about the first rate cut in September could shift to 25 bps v 50 bps if the next month of labor-market data is as crummy as Friday’s payroll report but rate cuts in between Fed meetings are very rare wsj.com/economy/central-bank…
Andrew Opdyke retweeted
Excluding government, education & healthcare (mostly government) and leisure & hospitality (still up from COVID lockdowns), jobs were up only 17,000 in July.
ftportfolios.com/blogs/EconB…
Andrew Opdyke retweeted
For the record. Paul Volcker did not stop inflation by raising interest rates. He stopped inflation by slowing money supply growth. The result of this was higher interest rates. It was the slower money growth that fixed inflation, not the higher rates.
Andrew Opdyke retweeted
The difference between the men's 100m freestyle final 88 years apart. 😮
Los Angeles 1932 & Tokyo 2020.
























