Pinned Tweet
Deficient test procedures and protocols is a very common Root Cause.
Experienced this several times on my projects at $CMI $CAT $GE
Here is an synopisis of corrective actions for this incident:
Based on the NTSB investigation report (Pages 14–15), the corrective and safety actions implemented by both Southwest Airlines (SWA) and Boeing addressed the maintenance testing process, high-wind monitoring, and inspection thresholds:
Southwest Airlines (SWA)
Overhauled the Standby Hydraulic Operational Test Procedures: Addressed the single-line requirement used during the "A Core Check" by replacing it with a dedicated work instruction card. This card provides a step-by-step procedure for the test and requires individual sign-offs by assigned mechanics (preventing missed steps, such as visually verifying rudder surface deflection at the aft of the airplane).
Fleet-Wide Inspection Campaign: Conducted general and detailed visual inspections across all 737 MAX aircraft targeting the vertical stabilizer at the main and standby rudder Power Control Unit (PCU) mounting and interface surfaces (no anomalies were found).
High-Wind Event Notification Process: Implemented an internal process to actively notify maintenance control whenever parked aircraft experience extreme high-wind events that may trigger conditional inspections.
Boeing
Established Defined Wind Thresholds in the AMM: Revised the parked/moored extreme high wind conditional inspection in the Boeing 737 Aircraft Maintenance Manual (AMM) to include a defined trigger threshold of 75 mph (65 knots). Previously, the manual lacked any numerical wind speed criterion, relying only vague qualitative indicators (such as aircraft jumping chocks, microbursts, tornadoes, or ground skid marks).
NTSB issues final report on the May 2024 rudder control system anomaly on Southwest Airlines (SWA) flight 746 during flight between Phoenix Sky Harbor International Airport (PHX) and Metro Oakland International Airport (OAK). web.ntsb.gov/investigations/…
ALT Damage to the standby rudder PCU support bracket and ribs of the vertical stabilizer. (Source: Southwest Airlines)
Shaker Cherukuri retweeted
Profit isn't the only signal. Feeling great about helping people. Changing people's lives. Helping your country. Putting people in a position to succeed. To enjoy their lives. People get to pick their own motivations. If they are fortunate, they will reach those goals.
That's the beauty of capitalism. People get to choose the reward that motivates them.
Entrepreneurs can choose to make as much money as possible. Or not. They can choose to make enough money to focus on using their most valuable asset, their time, to not have to work at all. They can choose to make money to help people.
But. Every entrepreneur that wants to reach their goals needs people. People to work for them, buy from them, sell to them. Even the person in the basement with nothing but AI, needs to connect to humans to reach their personal goals.
Capitalism works. But not for everyone.
Where Capitalism fails people, Capitalism isn't the problem. It's the people that put profits above all else that are the problem.
If you put your personal net worth or cash flow above all else, you may make insane amounts of money, but you will suffer the greatest indignity. You will be hated. Your time will not be your own. You will never experience the joy that comes from helping others.
No one is born knowing where their life will take them. We get to do our best, in our circumstances, to try to enjoy our lives.
Capitalism allows us to try to get there. Capitalism allows people who have achieved those goals to help others. Capitalism allows us to choose our own personal path.
Unfortunately, we don't all get there. But we get to try to choose the path we want.
The signal is personal choice. Not profit.
Mark. Do you hear yourself?
You say government doesn't have to make a profit, and that is true. It is the whole problem. Profit is the signal that tells a producer he created more value than he consumed. Government has no such signal, which is why it cannot know whether anything it does is worth doing. That is not a friction AI removes. It is the absence of the one piece of information AI would need.
And everyone else in your plan does have to make a profit. The researcher who spends ten years on a molecule that may never get through the FDA. The engineer who designs the scanner. The company that risks capital on a manufacturing line. The doctor who borrowed three hundred thousand dollars for the right to work ninety hour weeks. The nurse. The hospital that has to keep the lights on. Every one of them has to take in more than they put out or they stop, and when they stop, the care stops with them. Government not needing a profit means nothing if the people who make the thing still do.
Then there are the people whose money you intend to run through the Treasury. They earned it. They need it. And they have a right to it that does not depend on whether your commission has a good use in mind. A government that doesn't have to make a profit is a government that only has to take, and the taking is the part you keep leaving out of the pitch.
You know all of this. You built Cost Plus on exactly these premises. Publish a price, earn a margin, let the customer decide. You just refuse to grant the people who make medicine the same thing you granted yourself.
Shaker Cherukuri retweeted
The internet democratized information.
AI is democratizing the ability to use information.
That’s a much bigger revolution.
Having access to a million research papers means very little if you can’t understand them.
Having access to an intelligence that can explain, compare, analyze, and help you apply those papers changes everything.
The internet gave everyone a library.
AI is giving everyone something resembling a research team.
And we’re only getting started.
Shaker Cherukuri retweeted
Fortunately, I have 35 years of experience in biomedical research (anyone is welcome to check my scientific record), and I have been saying the same thing for years. I completely agree with Demis: we will cure all diseases within 10 years AND reverse aging in approximately 15 years!
I realize I am an exception among my colleagues, but what I find truly messed up is denying people, whether sick or healthy, this very real hope for the future.
Let me explain the main reason why even brilliant scientists and experts struggle to imagine how this could possibly happen. It comes down to extreme cognitive dissonance and the enormous ego of human intelligence: if WE cannot achieve something, we find it inconceivable that MACHINE intelligence could. In fact, the smarter people are, the harder it often becomes for them to accept this possibility, and the more elaborate their coping mechanisms become.
Top mathematicians have just lived through this realization, and physicists will be experiencing it soon.
Yet biologists somehow feel more secure. Not because they are smarter, but because biological systems are vastly more complex. After centuries of research by some of the greatest minds in human history, the idea that AI could solve problems in biology that have eluded us for so long, and do so within just a decade or two, seems utterly inconceivable to them.
Well, they will find out soon enough!
As someone who understands biology deeply, I know that its complexity is not insurmountable. Yes biology IS extraordinarily complex, but it is ultimately a system governed by physical laws and principles that can be decoded, which we have been doing albeit very slowly. Achieving this in a decade will require superintelligence, vastly greater computational power, and enormous amounts of biological data. That is why I believe it will take a decade or two, compared with mathematics and physics, where AI is already making extraordinary progress at a much faster pace.
The fundamental mistake is assuming that the limits of human intelligence are also the limits of intelligence itself. We have spent centuries trying to understand biology with human brains, same substrate it is made of. We are about to bring an entirely different level of intelligence to the problem, one that will eventually surpass the collective intellectual capabilities of all scientists who have ever lived.
As such, I stand 100% behind Demis and others, including Dario Amodei, who have made similar predictions.
Mark my words: the coming revolution in biology and medicine will be unlike anything humanity has ever experienced!
When Demis Hassabis talks about curing all diseases within 10 years, @dawallach says scientists who have spent decades studying these problems aren't necessarily buying it.
"When [scientists] hear AI people say smart things, they go, 'Cool, yeah, let's run it.' Everyone respects Demis."
"But when Demis says we're gonna cure all diseases in 10 years, I think people who have two PhDs and have been doing this for 40 years go, 'Come on, man.'"
"You're giving a sort of false hope to a lot of people who are legitimately out there suffering or have a loved one who's on death's doorstep. That's a little messed up."
Shaker Cherukuri retweeted
Imagine telling someone in 1850 that you’re opposed to machines replacing human labor because people need jobs.
They’d probably think you were insane.
People weren’t working 14-hour days in factories because they loved employment.
They were working because they had no alternative.
Yet here we are, nearly 200 years later, treating the preservation of human labor as some kind of moral obligation.
The objective was never to create jobs.
The objective was to create prosperity.
Somewhere along the way, we confused the two.
Here is what I did with my annual physical this year (2026):
1) Tests ordered in advance before the visit. This is a new protocol from the physician practice this year. I can see why this makes sense since you can discuss the findings during the visit instead of afterwords over the phone. Downside is may need additional tests and visit to the lab again. Acceptable trade off.
2) I used my AI assistant to analyze the results and compare it to results from prior year
3) Crated a document that summarized the results and comparative analysis into talking points
4) Came up with plan of action and recommendations for discussion
5) Discussed the plan with the physician and was commended on my preparation. He was very receptive and even asked to keep the printout I handed out. I suspect he might start using this protocol for other patients!
Win Win Win?👍
I want AI to replace my doctor.
Not because my doctor is bad. Because my doctor is human.
Tired. Rushed. 12 minutes per patient. Hasn’t read a new paper since 2014. Forgets what I said last time.
An AI that’s read every medical paper ever written, never gets tired, and has all the time in the world for me?
Yes please.
People call that dystopian.
I call it the first time healthcare has actually been about the patient.
Boeing 737 Max MCAS Root Causes were a combination of:
1) design flaws driven by
2) flawed requirements (IMO)
3) Flawed FMEA (Failure Mode and Effects Analysis)
4) AOA Sensor failure (2 crashes had different failure modes)
5) Field service operations and maintenance issues
6) Pilot training deficiency with MCAS
AI/SI review of the code and using AI/SI for code would not have caught all these issues.
This is the reason why physical systems that require robust fail safe fault tolerant system design engineering still require experienced system design engineers in the loop for the foreseeable future.
Use the AI tools as a tool. Use experienced system design engineers in the loop and for validation for sure.
FMEA is very crucial. Don't simply go through the motions to get past the design review.
Can the Super Intelligent systems get there eventually once they are trained on this kind of proprietary data hidden in the bowls of industrial physical systems?
What do you think?
Shaker Cherukuri retweeted
Speculation: Valeo’s new BlackBerry $BB QNX smart-camera wins in China may point to a more chip-flexible platform—and possibly a shift from Mobileye EyeQ on these programs. Could Eclipse S-CORE play a role? No replacement or S-CORE link has been confirmed.
Valeo/Mobileye: valeo.com/en/valeo-has-produ…
Valeo/QNX: stocktitan.net/news/BB/valeo…
~REALALOY
Shaker Cherukuri retweeted
I claimed that about 2–3 years ago, we entered a new era of science that I called Science 3.0, in which human scientists and AI work together to conduct research and make discoveries.
Science 1.0 was the era of individual scientists making discoveries, dating back to the beginnings of modern science. Science 2.0 was the era of teams of scientists working together, which became the dominant model of scientific research over the past century and almost exclusively so in recent decades.
Soon, within just a few years, we will enter Science 4.0, when AI, or rather superintelligence (SI), will make most scientific discoveries largely on its own. However, we humans will still be in the driver’s seat, directing which problems to solve, setting research priorities, and applying and validating discoveries in our wet labs.
After this relatively brief period of Science 4.0, perhaps lasting a decade or two, we will enter the final phase of scientific research and discovery: Science 5.0. At that point, superintelligence and autonomous robots will independently drive and determine virtually all scientific progress. The intelligence gap between us and SI will have become so vast that our own intelligence will be too limited even to formulate the right questions, let alone comprehend the full depth of the answers.
In fact, I recently told a friend and fellow scientist that we, the scientists trained today, are the last generation of human scientists. I believe this will turn out to be one of the most profound transformations in the history of humanity.
I feel incredibly honored and privileged to be part of this extraordinary transition, and deeply grateful to be alive to witness humanity passing the torch of scientific discovery to a far greater intelligence.
For thousands of years, we have struggled to unravel the mysteries of the universe with the limited intelligence evolution gave us. Soon, we will pass that quest to minds far more capable than our own, minds that may finally uncover the deepest secrets of reality, life, consciousness, and perhaps even the ultimate meaning of our existence.
What a magnificent time to be alive, to witness the final chapter of human-led science and the beginning of an age of discovery beyond anything we can possibly imagine!
i don't think people realize how quickly we're getting to the point where AI can do actual scientific research basically without humans
yesterday Nature published a paper about an AI-run laboratory in China that autonomously does DNA synthesis, amplification, sequencing, among other experiments. more than 11,000 experimental steps with over 20 different instruments
the humans are basically there to refill the reagents and maintain the equipment
at the same time Anthropic published research where Claude helped a team discover and determine the structure of a bacterial-virus vault particle, searching through 15 BILLION protein sequences and doing the computational analysis
just think about where this is going
AI comes up with an experiment, robots run it, instruments return the results, AI interprets them and proposes the next experiment. repeat
once that loop becomes reliable enough and generalizable, scientific progress is no longer bottlenecked by how much research humans can personally do
scientific discovery is becoming scalable
Tesla Will most likely exit the automotive vehicle sales to consumers business model after merger with SpaceX in 2027/28.
Focus on robotics AI/SI, energy, logistics/transport etc. higher margin businesses.
