
Assoc Prof @PittLaw & @SciPitt, fellow @yaleisp, 🐕s🐇s, 🏀⚾🏈, @Cal alum #cybersecurity | #blockchain | #law | #techpolicy (RT,♥️ ≠ agreement, views mine.)
Dallas, TX
Joined February 2009
- Tweets8.2K
- Following916
- Followers1K
- Likes16.1K
David Thaw retweeted
If you work in tech policy long enough, you'll see lots of friends sticking theory-based views as techincal & political realities change. Especially true of academics, who stick to outmoded ideas for decades! It's quite jarring for pragmatic folks who are trying to make a change.
David Thaw retweeted
A lot of places learning at a visceral level the meaning of key man risk this week. All the committees and PowerPoint decks and placemats in the world cannot paper over the reality of the unique mind doing things that others can barely explain in hindsight.
David Thaw retweeted
1. You can tell me it's a waste of time for schools to teach literature, but half the executives in Silicon Valley seem to have misunderstood their favorite book, with potentially devastating consequences.
David Thaw retweeted
Be the teacher and scholar you trained years to become.
See our open searches: uaustin.org/careers
David Thaw retweeted
We are hiring an Open-Rank Professor of Economics at @uaustinorg. Start Fall 2027, priority review for applications by Jan 4, 2027.
Please share with any candidates that would be a good fit!
jobs.lever.co/uaustin/1e6d12…
David Thaw retweeted
12 unsolicited emails *today* from high schoolers, undergrads etc., 8 of them fully AI-generated.
I used to read and respond to these, I just can't anymore. The cold email used to be a positive signal of confidence, entrepreneurship, willingness to risk rejection, having done a bit of research. Not anymore.
Capable and entrepreneurial people will need to find a new way to distinguish themselves. I'm not sure what it will be. Maybe by building something original.
It's a lot harder than sending a cold email, but that's good, it means there's still some signal in it.
David Thaw retweeted
Y2K catastrophism tracing a direct psychological line to AI/SI doomerism hits the cyber intelligence profession harder than would be typically credited. Much of the earliest commercial cyber intel coverage from shops like iDefense went hard on projections of fears but too often untempered by continuity & resilience net assessment. The intel adjacent analysis supporting much of the worst case scenario arguments over hyperscale compute outcomes is of a piece. But there are real matters of concern here, demanding serious consideration and clear eyed methodological approaches.
David Thaw retweeted
“The Chronicle’s freelancer policy asks that submissions be ‘written by you, not AI’ (with rare exceptions). I wrote this essay the way I wrote the article that inspired it, and I invite its editors and readers to judge it accordingly.”
“The author drafted this essay in extended collaboration with Claude Fable 5.1.”
The story of “the article that inspired it”:
“A professional journal recently accepted an article of mine. The editor congratulated me on the acceptance and then made an unusual request. Could I submit another draft written ‘in my voice,’ since this one had some sentences that ‘read like something written by AI’? If I preferred, the editors would revise the piece themselves, ostensibly to ‘humanize’ it, and I could approve their changes.”
“This amusing episode helped me to understand a confusion that now runs through the academy. We have begun to judge analytical writing not by its content but by its suspected origin, which no reader can verify. Because an analytical essay is an argument, and an argument can be evaluated, professional analytical writing should be judged as it always has been: by its quality.”
David Thaw retweeted
Survey of 1,059 UC faculty finds 86% support reinstatement of SAT/ACT requirement ucstudentsuccess.org/wp-cont…
David Thaw retweeted
*Mathematicians did not ask for this work to be done"
WTF?!! Since when was math something that was only done at the request of professional mathematicians?!!
Mathematicians don't own mathematics! This is crazy!!
David Thaw retweeted
Who will be dean of UATX Law? The search has begun.
Search committee: Judge James C. Ho, Judge Andrew S. Oldham, Judge Amul R. Thapar, and co-chairs Claire Brickman and Adam Biggs.
Applications reviewed beginning October 20. Know the right person? Send this to them.
Apply here: jobs.lever.co/uaustin/540fe3…
David Thaw retweeted
I completely disagree with this. Claude Code has me thinking harder than any job I've ever had.
But I get why Braeden feels his brain melting. The cause is right there in his own post.
Look at how he describes the work. Five or six terminals open for six months, waiting on responses just to smash enter 90% of the time. His words. He calls that the whole job now.
If pressing enter is 90% of your job, the model is making your decisions and you've become the keyboard. He automated the judgment along with the typing, and judgment is the part that keeps you sharp.
There's actual research on this. MIT's Media Lab put EEG caps on 54 people writing essays, and the group leaning on ChatGPT showed the weakest brain connectivity in the study. Most of them couldn't quote a single sentence from the essay they had just turned in.
The part of that study almost nobody mentions is the crossover. People who wrote with their own brain first, then brought the model in afterward, measured higher engagement than everyone else. Same tool. The difference was who did the thinking first.
I run my whole content business on Claude Code with a small team, and the days drain me more than any PM job ever did. The exhaustion comes from judgment. Looking at a dozen working prototypes and deciding which one deserves to exist, or realizing the spec I wrote at 9am produced the wrong product by 11am. Mostly it comes from saying no, over and over, because when building is free, shipping everything buries the three things that matter.
The people who feel sharp after a day of this are exhausted from decisions. The people who feel dull are exhausted from waiting.
So if Claude Code has you smashing enter all day, change how you drive it:
1. The full Claude Code tutorial for PMs: youtube.com/watch?v=4nthc76r…
2. How to use Claude Code like a pro: news.aakashg.com/p/how-to-us…
3. Build your PM operating system: news.aakashg.com/p/pm-os?utm…
4. Context engineering, so the model runs on your judgment: news.aakashg.com/p/context-e…
The builders who come out of this era sharper will be the ones who kept the judgment for themselves and handed the model everything else.
RT @OrinKerr: A lot of people respond that using AI agents is uniquely dangerous, and that this justifies extra broad criminal laws on the…
This quoted post is unavailable.
A lot of people replied to this tweet by saying that I'm just wrong about criminal law: We *do* normally make it a crime to commit tort law negligence, as shown by statutes like negligent homicide.
Two thoughts.
First, it's true that there are exceptions to the ordinary rule that negligent acts are not made criminal—most prominently, negligent homicide. But the word "negligence" is generally used in criminal laws like negligent homicide statutes differently than it is used in tort law. "Negligence" is not tort negligence; it's a higher standard more like gross negligence. See e.g. People v. Penny, 285 P.2d 926 (Cal. 1955).
Using the Model Penal Code's framework, which courts have in the CFAA setting, the "negligence" in criminal laws like negligent homicide statutes is focused not the failure to meet a reasonable standard of care like you have in tort negligence. but instead on an egregious failure to even realize one's acts are substantially likely to directly cause the result, such as death. It's not the same thing as tort law negligence, although it uses the same word.
Second, and more broadly, negligent homicide liability is thought to be a special case. It's premised on the understanding that conduct that proximately causes a significant risk of loss of human lives is a unique type of harm, and that a general law dealing with causing that loss—in any way, so long as it proximately causes the significant risk of loss of human lives and then human lives are lost—is needed.
Some say, ok, but we need to apply that law to death caused by AI agents. We shouldn't exempt death by AI agents from the existing legal framework for negligent homicide. But we already do that. Negligent homicide laws already cover causing death through AI agents, as they cover causing death in any way. They are result crimes, not means-of-conduct crimes, so all that matters is proximiately causing the result.
I think the idea of having a federal crime for causing an act of unauthorized access through an ordinary tort standard of negligence is really different.
This doesn't mean you couldn't have a new law. If a farmer designs a bad fence and his sheep get out and damage a neighbor's property, my understanding is that we hold the farmer strictly liable, without a negligence standard. It would make sense to do the same for AI agent harms.
Also, there are interesting questions about whether a reckless mens rea (in the MPC framework, again) could make sense as a standard for criminal liability. More on that in a post next week, I hope.
David Thaw retweeted
In 1970, an economist wrote 13 pages explaining why good used cars vanish from the market. Three journals rejected the paper as trivial. It won the Nobel Prize, and it just predicted the death of the compliment.
George Akerlof's insight was simple. Sellers know if a car is a lemon. Buyers don't. So buyers only ever offer the average price, anyone with a genuinely good car refuses to sell at that price and walks, and the market slowly fills with junk until buyers lower their offers again. The spiral runs until only lemons trade.
The engine of the whole collapse is unverifiable quality.
Now look at what "is this AI?" actually is. Praise was the currency creators got paid in. "So well written" really meant "I can see the hours you spent." The compliment was payment for verified effort.
AI made effort unverifiable. A stunning illustration might represent 40 hours or 40 seconds, and you can't tell which from your phone. So the rational reader does exactly what Akerlof's car buyer does. He stops paying full price to anyone and pays the average instead, and the average assumption online in 2026 is "probably machine."
Then the spiral kicks in. Human artists bear real costs, so lemon prices push them out first, which raises the share of generated content, which makes the skepticism more accurate, which drives out more humans.
Akerlof needed used cars to prove the theorem. We're running it on everything people make.
His paper's fix wasn't better cars either. Warranties, brands, dealer reputations, mechanisms that make quality verifiable again. Which is why the next decade of the creator economy will be a war over proof of process rather than proof of talent.
David Thaw retweeted
David Autor é o economista do trabalho mais citado do mundo. MIT. Ele e o Google fizeram algo raro: um experimento científico controlado (RCT) sobre IA no trabalho.
O que é RCT: você sorteia quem recebe o tratamento e quem não, compara os dois grupos, e mede a diferença. Padrão-ouro da ciência. Aqui o "tratamento" era uma IA pra redigir patentes.
133 advogados de 11 escritórios americanos, divididos por sorteio. 2/3 usaram IA por 3 meses, 1/3 não. Produção avaliada às cegas por outros advogados especializados.
Com IA ligada, todo mundo melhorou. Qualidade dos juniors subiu quase o triplo da dos seniors.
Aí tiraram a IA e pediram pra todos revisarem um documento sem nenhuma assistência. Seniors que tinham usado IA ficaram melhores que quem nunca usou (p = 0,02). Juniors? Ganho médio zero. As notas deles se bifurcaram: menos medíocres, mas mais ruins E mais bons ao mesmo tempo. Pra uns a IA acelerou aprendizado, pra outros substituiu.
O senior já tinha construído julgamento crítico ao longo de anos. Quando usava IA, conseguia avaliar as sugestões e internalizar padrões. O junior pulou essa construção. Delegou o raciocínio pra máquina. Enquanto tava ligada, funcionava. Desligou, ficou sem base.
Tipo calculadora: se você já sabe aritmética, ela te acelera. Se não sabe, você não funciona sem ela.
Expertise prévia é pré-requisito pra extrair aprendizado durável de IA. Quem mais se beneficiou foi quem menos aprendeu.
fonte: NBER Working Paper 35720, Autor et al. (set/2026)
David Thaw retweeted
25 years. A very different young man, sitting about a week into new assignment in Geneva, when the world changed. Not knowing yet who we had lost, and who lived. Nor who else would fall in the decades of the Long Wars to follow. Scrambling to exfil others still forward deployed, unclear where or when this wave of hits would stop. Only the dim growing awareness of the hard things we would have to do next. With no concept of how long it would take, or how these things would evolve.
There is so much I might have said to my younger self in the intervening years.
Today I only admonish all (including that young officer or analyst and every other like him later) to never forget the first purpose of intelligence, to provide warning so as to avoid surprise. Our modern professional history began at Pearl Harbor. But it was never about the planes.
David Thaw retweeted
Wow…Roland Fryer and @BillAckman have created a $1 million prize for academic truth-telling, the Carob Trust Prize for Academic Courage.
“The prize awards $1 million each to as many as five social scientists a year who have demonstrated intellectual independence, published findings that were attacked rather than answered, and been validated by the evidence—despite the professional cost. The selection criteria are designed to distinguish courage from contrarianism: Nominees must show a sustained commitment to following logic and evidence regardless of pressure, a willingness to ask questions others avoid, and work that has shifted academic debate, public discourse or policy—often despite being misread, mischaracterized or vilified at the time of publication.”
David Thaw retweeted
How is this message supposed to make the younger generation mentally healthier or less depressed? All we're doing is whispering into their ears: “You know what? You can't handle even the most basic challenges that generations before you handled just fine. You will crumble, you will fall, you will be permanently injured. You're so depressed and anxious now, so we need to protect your oh-so-fragile, constantly endangered self from anything that might give you a sense of meaning and accomplishment.” It may make the protector feel virtuous, but it certainly harms the recipient.





























