Patrick Skinner - edu/acc

For those who are interested in working in EdTech, I need to be honest with you about what you're signing up for. You probably shouldn't apply. EdTech's graveyard is full of brilliant engineers. AltSchool raised $200M and closed. Knewton raised $180M and sold for scraps. 2,148 edtech startups in India shut down in the past five years. They didn't fail because of a lack of technical skill. They failed from mission drift. Here's what working in edtech actually means: You're not building software. You're building a theory of how children learn. - @MitchForest When you're 2 months into the same feature, and it still doesn't work, you won't have the dopamine hits of shipping fast. You'll need something deeper. For the engineers who build learning apps for @AlphaSchoolATX, they are REQUIRED to spend 2 hours every day studying. Not coding. Studying and reading papers on cognitive load theory, motivation research, and translating learning science into applications. Two hours. Every single day. Engineers like @yiran__c write about the neuroscience of handwriting. @arpangup shares the Harada Method. @LamarDealMaker explores AI for skill development. This isn't optional. It's the job. Your success metrics will lie to you. High DAUs don't mean kids are learning. Viral engagement doesn't mean mastery. The metrics that drive consumer tech success actively undermine educational outcomes. The money will tempt you to build the wrong thing. EdTech market: $400B by 2030. That capital creates pressure to optimize for pitch decks instead of classrooms. Byju's hit $22B before collapsing. An MBA analyzes constraints. A builder changes reality. What we require: ➡️ Mission Alignment ➡️ Agency ➡️ Strong Engineering Skills If you're chasing market opportunity or building for your resume, don't apply. But if you're a builder who sees broken systems and feels an overwhelming need to fix them? If you believe high standards create happy kids? If you want to change education for a billion children? Before applying, read the link in the reply. And if you're still interested after understanding what it really takes, let's talk about joining us in building the future of EdTech.

How I look while coding.

Go's Edge: - Error handling is extremely simple. - No classes, decorators, or type gymnastics to overthink. • AI nails the goroutines and channels pattern every time. Simple language = predictable AI output. TypeScript's: - AI is heavily trained on it. - Full-stack/frontend tasks are almost always TS - Complex type inference More training data = smarter completions. Openclaw has some UIs, but TS is an extremely strong frontend while supporting backend/automations. Go is not intended for the frontend. It's the go-to language for things like Backend APIs and microservices, which is a large amount of what OpenClaw does.

@PSkinnerTech @JacobESalazarS Better than TypeScript?

Going to test all the OpenClaw alternatives today to figure out where they shine and if there’s anything net better yet (entirely possible.) What else should I test? Will post results to subscribers to pay for compute x.com/Austen/creator…

Applications are open for Cohort 5 at Gauntlet AI. The countdown is on. Cohort 5 starts April 27. Learn More → gauntletai.com/apply

Polymaths in this era will be undefeated

@eduleadership "I’m not saying it can never work, but it would be truly extraordinary..." There's a really great Edison quote about this...

The problem is that a lot of smart people are working on a solution category that has been proven not to work. I’m not saying it can never work, but it would be truly extraordinary, and people don’t seem to have any particularly extraordinary ideas.

@eduleadership So, you think that instead of becoming part of the solution, because you have no solution, you'd rather hate on the others who are trying to develop a solution. Because there's clearly a problem, and many are trying to develop a solution... except you.

@PSkinnerTech Ha, this is not my job or my field of research. I’m simply pointing out the research of others, which is widely available. I don’t have my own methods. I’m simply pointing out that EdTech is worse than the low-tech status quo.

0.15 SD on a course final exam means nothing. Noise.

@JacobESalazarS @Austen AI is extremely good at writing in Go. I chose it almost exclusively based on that alone.

@PSkinnerTech @Austen Are you particularly interested in ultra-lightweight ? I’m pretty happy with nano law, or is there any additional benefit I’m missing ?

Then why don't you become part of a genuine study and introduce some of the best edtech into a learning environment to test it for yourself? At this point, you're a professional hater with zero data on your own end to refute whether your methods are genuinely better. Unless I'm clearly missing something.

@PSkinnerTech I believe that claims about measurable + effects on student learning are mostly bogus. Large negative effects are by far the easiest to detect and avoid, so I consider criticism to be valuable work.

@eduleadership Every single time you scream "selection effects," do you have proof that's what's happening within that study, or are you just assuming that any EdTech that is reporting high increases of learning outcomes is only due to selection effects?

@PSkinnerTech I’m not sure what you’re asking.

@eduleadership I would agree with that.

@PSkinnerTech I mean, if I divided my class into two groups, applied a treatment to one, and that group scored a single point higher on the final exam, I definitely would not publish an academic paper about it.

@JohnDMacari @TheFinestCast That outside foot trip was beautiful.

NYPD Police Officer Grant Pulagrin is my pick for 2026’s “Cop of The Year” Let’s talk about the reality of what officers like Grant Pulagrin face each day using this video as an example. In the below video you witness Grant make a beautiful solo, open field tackle and apprehend a violent terrorist who just threw 2 IED’s a protest. In that moment, it was completely reasonable for him to believe the suspect could still be carrying another explosive. Despite that risk, Pulagrin takes the suspect to the ground, mounts him, and delivers strikes to gain compliance ending the threat and getting the terrorist into handcuffs. That’s policing, that’s courage and that’s what we pay, train and expect police officers to do. But here’s the unfortunate reality for NYPD cops in 2026: While every reasonable person knows this was a reasonable and justified use of force it’s important to understand that his actions on that day generated a use of force report, which generate an automatic notification to NYPD Internal Affairs and NYC’s Civilian Complaint Review Board (CCRB) to investigate that use of force. Grant is now being investigated for his use of force by both the NYPD and the CCRB. Due to the media and public attention on this case it is reasonable to believe that those investigations will clear Grant’s use of force but regardless of the outcome, these self generated allegations remain on his record forever, in retirement and beyond. If the suspect or his attorney files additional complaints as a legal tactic, which also happens most of the time, those allegations also live on permanently. The mount of the terrorist is also a move that was outlawed by NYC’s Council as part of the diaphragm law, which outlawed not only chokeholds as but also outlawed putting pressure on the back or chest of suspect. Again because of the attention around this case it’s unlikely Manhattan DA would bring charges against Officer Pulagrin but the potential to do so is there. Now imagine this wasn’t a terrorist attack. No media headlines but the same exact scenario plays out and Grant reacts the same way and does just as an amazing job apprehending a fleeing violent suspect. In this scenario Grant more than likely receives a substantiated CCRB as a result of their “investigation” into his self generated allegations. This scenario would cause Grant to not receive qualified immunity and not be indemnified by NYC, meaning Grant would be held personally liable for damages from a lawsuit. Grant would also have his professional record marred for life and could face a possible criminal indictment, for doing HIS JOB ! So when the headlines fade, don’t forget what you saw. A police officer ran toward danger, took down a terrorist, and protected innocent people while knowing the system he works in may still come after him for doing his job. Next time you judge a cop based on their CCRB or lawsuit record remember that. GOD Bless Police Officer Grant Pulagrin ! GOD Bless The NYPD ! FYI @nypdpc hopefully it’s Detective Pulagrin sooner than later.

@eduleadership Just to make it clear, you have no measurable effects, and you're just critical of everyone else's because you believe good results are overwhelmingly due to selection effects. Is this correct?

@PSkinnerTech If you want my actual quantitative estimate, it’s that teachers impact about 10% of the variance in student learning, and administrators about 1%. EdTech is more analogous to a book than a teacher, so I’m not sure any such estimate would be meaningful, but it’s ~0 at scale.

@eduleadership Do you have tangible data that shows that's what's happening every single time? Or is that simply your default assumption?

@PSkinnerTech To understand my antagonism toward EdTech, it’s important to understand that virtually all apparent results in education are, overwhelmingly, selection effects. Pick kids who will do well, and take credit. 🥳😅

@eduleadership Maybe adding the "if this is just noise" tie too much to a single point of data, rather than your overall views of EdTech. I agree that .15SD could have many different factors involved and is marginal at best.

@PSkinnerTech If you’re talking about personally…teaching a lesson well. Reviewing. Encouraging students to do their best on a test. 0.15 SD is probably 1-2 points on the final. Odd that people want to claim that as some sort of moral victory for AI.

@garrytan "But it's just a repo full of prompts, Garry!"

Many such cases

Southern states boost early reading, but gains stall in middle school. Why the ‘Mississippi miracle’ disappears by eighth grade. hechingerreport.org/proof-points-8…

Hey @NaniSkinner, I found a new project for our kids. @chiefhype, @CameronSorsby, @nateliason, thoughts on ideas like this? I'm happy to help.

Openclaw has completely unlocked the idea of developing a steampunk-themed STEM club where we rebuild everyday tech from the ground up. Cell phones, smart watches, etc.