R.Dal (@ryandal) header
R.Dal (@ryandal) profile picture

International school leadership. On holiday. Probably wrong.

South East Asia
Joined February 2009
Photo by R.Dal (@ryandal)Photo by R.Dal (@ryandal)Photo by R.Dal (@ryandal)Photo by R.Dal (@ryandal)Photo by R.Dal (@ryandal)Photo by R.Dal (@ryandal)Photo by R.Dal (@ryandal)Photo by R.Dal (@ryandal)Photo by R.Dal (@ryandal)Photo by R.Dal (@ryandal)
What is the planning fallacy and why is it a potential problem for school leaders? open.substack.com/pub/pendid…
1
9
1
23
I traveled 32,649km for 28 days across 3 countries to explore the most ambitious projects reinventing education. We are filming the world’s first documentary covering this new generation of learning models. Along the way, I met with 32 founders, teachers, parents, students, and experts, and got to explore 13 projects covering everything from AI tutoring, microschooling, and precision learning to the Socratic method, project-based learning, homeschooling, new VC models, and school choice programs. It’s been an intense but incredibly rewarding journey. I was blown away by the results of these projects and the passion of the people behind them. We’re witnessing a renaissance in education, a Cambrian explosion of new learning models. And it’s just getting started. Legacy education is cooked. The filming phase was a massive success. Now it’s time to bring all these stories together and make an amazing movie for y’all. Many of you have been asking how to support the project. Now you finally can. You can grab a ticket for the exclusive global premiere in April 2027, become a Screening Host and organize a local screening, or become a Funding Partner and get an exclusive video on the Ipe Decode series. 🔗Buy tickets and support the project at networksociety.film/episode3 LFG!
19
20
7
242
21,958
Thrilled to announce $5B in funding for @IsomorphicLabs to accelerate our mission to Solve All Disease. We continue to make rapid progress reimagining drug design from first principles with AI to radically cut the time & cost to find new medicines. Gradually, then all at once...
616
1,526
402
14,213
913,258
One of the coolest things I've ever seen!! Loved the thunderclap as the starship dropped out of the heavens, before the catch. True alien tech. Huge congrats @elonmusk and the @SpaceX team! the whole stack is now proven out. Ad Astra! 🚀
242
746
71
12,615
1,640,779
Everyone is all against alcohol and sugar now. But who was there when your life was falling apart? It sure wasn’t fucking salad.
641
10,722
711
80,580
2,650,892
All the best people declare war on Hogwarts. Have you ever considered dividing your gender identity into separate pieces and storing it in hidden objects, so you can’t be invalidated to death?
2,308
6,088
776
91,539
3,205,167
Microsoft's diagnostic AI solved 83% of 304 complex New England Journal of Medicine cases. Generalist physicians with reference materials solved 36%. We need to get this technology into clinics safely and fast. We need to stop making people wait for the right answers.
148
439
88
3,669
305,178
I think lots of people are confused about this Alpha business. When you see someone who has advocated for a specific set of ideas for a long time go on to join or lead an organisation that seems completely antithetical to those ideas, then it's reasonable to be confused. Yesterday, I briefly discussed this with @PepsMccrea, and I argued that Alpha represents educational progressivism with a veneer of "learning science" (or what we have normally called "cognitive science" in England). Here, I'd like to expand a bit on that. I'm not going to cite every claim, because this has been part of common discourse here for at least a year (as long as Carl [and I think Peps]) have been working for Alpha. First, instrumentalism. In England, educational traditionalists (like Carl and Peps) argued that education and the education system is not designed merely to prepare students for the world of work. Instead, it is there to pass on cultural heritage, and knowledge for its own sake. Progressives on the other hand argued that it was more about "21st century skills" and railed against a system that "failed to prepare students for jobs that hadn't been invented yet." It's clear that Alpha is built in this vein. It runs courses on entrepreneurialism, media training, startup founding etc etc. NONE of their marketing that I have seen is about the joy of learning or being inculcated into a cultural or intellectual tradition. Second, long term learning. Proponents on this site have boasted about how students finish courses in a matter of days. One of the very first things I learnt from the traditionalist school was that learning is a process that occurs over the long term, and that short term performance means little. Third, "time back". Alpha uses an app called Timeback, which broadly - in my understanding - means that if students do well on their 2 hours of AI-enabled academics, they get more time to do other stuff. This again emphasises that academic education is something to be "got through" so we can get to the fun stuff: Navy SEAL training, social media, horse riding and...entrepreneurship. Fourth, in a similar vein, extrinsic motivation. Alpha has heavily pursued gamification and an extrinsic reward system. Again, English traditionalists used Self Determination Theory as a key plank of their arguments about schools and education, and extrinsic motivation as an educational lever has been widely derided based on simple findings from the SDT literature. Fifth, personalisation. Every so often someone here said we should teach shakespeare through rap or dance or whatever. Traditionalists thought this was awful. The content is the content, and it is our job to take students beyond their horizons, not pander to their narrow set of cultural experiences. And yet this is exactly what Alpha does. Writing lessons are based on students' particular interests, so Harry Potter and minecraft. Alpha students have spoken about how Romeo and Juliet etc is outdated and boring. Again, very much at odds with what proponents have said in the past. Sixth, trashing the traditional model. We suffered our way through Shift Happens and Schools on Trial (or whatever it was called) which went on and on about factory models and Gradgrind and sage on the stage etc. And yet Alpha promotional materials do the same. Seventh, children on social media. This was the most surprising thing this year, where students at Alpha schools were tasked with setting up an X account and getting a certain number of followers. This is obviously a safeguarding concern, and if an English school leader did this it is likely they would be out of a job pretty quickly. Again, opposing social media and teens on phones has been an important part of the traditionalist movement. Eighth, it's an experiment. This is a kind of trial - "we are trying to figure out if we can do this." There's a couple of problems here. First, children aren't guinea pigs (even if their parents are rich). Second, trialling it in the best possible circumstances (i.e. a rich private school) seems like a good idea because the risk is low, but it's also why it won't scale. Anyone who has tried to roll out educational change knows that there is an "adoption curve" and a chasm between your early adopters and everyone else. What works for one categorically doesn't work for the other. This is basically as close to an iron clad law of business and scaling tech as you can get. But also, if you really wanted to make change and improve the lives of children globally, you'd focus on the things we know work because we've seen them work! Everyone on that team knows about Michaela, Mercia, Ark King Solomon, West London Free School, Mossbourne, TTA and more. Schools that do incredible things for underprivileged students with a fraction of the budget. High academic expectations. Expert teachers. Brilliant behaviour. Knowledge-rich curriculums. If we really care about scaling educational solutions, let's start there. While we're at it, the focus of the traditional movement was always on disadvantaged students. What can we do for them? How can we help them? How do we serve those who need it the most? Alpha is so obviously not that. It's a hyper-expensive school for the children of elites. Of course, those students need educating too, but dedicating so much time and energy and intellectual capital to them seems at odds with everything that has come before. Ninth, and it's going to be hard to say this without being catty, but respect for academia and qualifications. I want to be super clear that I massively respect the intellectual acumen of people like Carl, Peps, Becky and others. But, they are not "learning scientists." To the best of my knowledge, nobody on the team has produced original research in that particular field. Again, that doesn't mean they are not worth listening to or whatever - but part of the traditional revolution in England was discovering these researchers who had made incredible findings about learning and the brain, within the confines of the peer-review and scientific inquiry method. I know that arguments from authority aren't empirical, but being able to rely on people like Kirschner, Willingham, Dunlosky and others was critical to the success of the traditional movement. Tenth, transferable skills. One of the founding principles of educational traditionalism is that you cannot teach generic or transferable skills like "Leadership" or "resilience", but the Alpha website and promotional materials are full of this stuff. Eleventh, track record. Alpha people have been talking about how fantastic their content is for at least a year, but we are now told that it is just "ok." Obviously most of it is proprietary, but someone posted some physics stuff last week and I thought it was pretty poor. Michael Pershan has blogged similarly, and Dan Meyer (!) has pointed to the objective data from a school that adopted Alpha and ended up doing worse. Twelfth, similar - ratification. Part of why England has succeeded is that we have objective metrics like standardised assessments and progress 8 that allow us to test the claims of schools and educationalists. Alpha has no such metrics, and instead we need to take their word for it that they are causing "2x learning." It's a surprise to see people who have been part of the positivist and outcomes-led policy movement wanting to be involved in this kind of non-verifiable model. Thirteenth, real books, but as a totem standing for traditionalism. A number of people have asked if Alpha students read real books. On paper. Great books. There hasn't been much of an answer, beyond a picture of a library. Ok, books are important, but they aren't really my point here. My point is that there is a romanticism around traditional education. Something that goes beyond the petty mechanics of pedagogy, and taps into something social and cultural about learning. That we are one link in a long chain of transmission, a chain which cherishes books not because of a fascination with bleached wood pulp, but because of what they represent: our ability to transcend the narrow confines of our lives and swim in the streams of collective human intellectual endeavour. Nothing I've seen at Alpha tells me there's anything like that idealism. Instead, it strikes me as a desiccated and hollowed-out vision of education, where students learn not from another human, but from an AI-generated screen. It doesn't surprise me that some people are interested in Alpha. It seems exciting and innovative as a business model. But it does surprise me that THESE people are interested in it. I harbour no ill will towards them, and as I've said I respect their work. I hope this comes across as respectful and measured. But it's no surprise that people are surprised.
39
75
32
419
74,202
So to her great credit did my mother, fortunately she tells me she laughed out loud at my gaffe for which I have apologised.
I had a little chuckle at the opening paragraph of Jacob Rees-Mogg's latest piece.
91
406
17
7,596
261,332
Meta-analysis of retrieval practice, spacing, and interleaving in classrooms concludes fairly robust average effects of each (open access) link.springer.com/article/10…
10
48
5
174
11,939
'The Archbishop of Canterbury… accused me of demoralising the English people. Sir Harold Wilson, a more powerful authority on ethics and morals, accused me of debauching the[m].’ - Harold Macmillan recalls the introduction of premium bonds, first sold on this day 70 years ago
22
109
23
785
189,886
This is a fantastic idea - a mass attempted replication of old findings of LLMs with new models. Much conventional wisdom is outdated. Recent models no longer invent citations and both hallucination rates & stereotyping has dropped. But they are still sycophantic & lean left.
236
159
44
1,697
108,775
Some have pointed out that AI fears help to sell AI equity & products, & raise the status of AI workers. Others then counter than those who fear seem to be sincere. But motivated bias is largely unconscious. Sincerity isn't much of a protection against bias.
Mary Phuong of DeepMind on AI loss of control risk. "The last 2 minutes gave me chills. A significant amount of people saying that AI companies are paying people to spread fear about these agents. I have never really felt like [loss of control] was an actual thing to care about. But seeing Mary and how her face trembles, made me reconsider my stand. She's either a Hollywood-level skilled actress, one of the best, or we should be scared. For real." "Seems like Mary is already aware of more than she can admit and it haunts her..."
17
8
1
115
10,841
I joined @GoogleDeepMind to develop a research program on the economics of AGI. Today, together with Andrew Koh (@andrewjkoh) and Shane Legg (@ShaneLegg), we are incredibly excited to release The Economics of Intelligence, the first in a series of projects from our group on how AGI will change the economy. How do capability improvements transform the economy, and when? In this project, we define thresholds of economic capabilities, e.g., minimal AGI, based on how they transform the economy. We then track how the economy–market structure, employment, and wages–changes during the transition across capability regimes. We have seen exponential increases in what AI agents can do but the broader economy seems almost remarkably unchanged. Part of the reason is that the solo AI agent is the wrong economic object; economic transformation is an equilibrium outcome where organizations compete. Looking at how intelligence functions within organizations, transformation will depend on when two key capability thresholds are crossed: competence (gains on individual skills, e.g., coding, math) and coordination (gains on ability of agents to cooperate and work together). We show that what economic transformation looks like depends on whether AI becomes supercompetent, supercooperative, or both—the regime of artificial superintelligence (ASI). Tracking this transition shows why the real visible transformation may come from improvements in coordination—a capability most benchmarks don’t yet track. Past a certain threshold, AI capabilities can break the Coasian ceiling on firm size and lead to an economy that looks drastically different than the one we see today. Two main takeaways: 1. To understand the path of transformation, we should be tracking agentic coordination more closely. 2. The impact of AI will likely show up in wages before market structure or employment. We are really excited about this paper. Excited to hear people’s thoughts. Link: institute.deepmind.com/essay…
248
784
210
4,861
547,745
This research gets at something missing from most debates about AI’s economic impact: coordination. Organizations are more than the sum of their best performers, so smarter individual agents won’t transform the economy on their own. Agents that can coordinate together might do.
I joined @GoogleDeepMind to develop a research program on the economics of AGI. Today, together with Andrew Koh (@andrewjkoh) and Shane Legg (@ShaneLegg), we are incredibly excited to release The Economics of Intelligence, the first in a series of projects from our group on how AGI will change the economy. How do capability improvements transform the economy, and when? In this project, we define thresholds of economic capabilities, e.g., minimal AGI, based on how they transform the economy. We then track how the economy–market structure, employment, and wages–changes during the transition across capability regimes. We have seen exponential increases in what AI agents can do but the broader economy seems almost remarkably unchanged. Part of the reason is that the solo AI agent is the wrong economic object; economic transformation is an equilibrium outcome where organizations compete. Looking at how intelligence functions within organizations, transformation will depend on when two key capability thresholds are crossed: competence (gains on individual skills, e.g., coding, math) and coordination (gains on ability of agents to cooperate and work together). We show that what economic transformation looks like depends on whether AI becomes supercompetent, supercooperative, or both—the regime of artificial superintelligence (ASI). Tracking this transition shows why the real visible transformation may come from improvements in coordination—a capability most benchmarks don’t yet track. Past a certain threshold, AI capabilities can break the Coasian ceiling on firm size and lead to an economy that looks drastically different than the one we see today. Two main takeaways: 1. To understand the path of transformation, we should be tracking agentic coordination more closely. 2. The impact of AI will likely show up in wages before market structure or employment. We are really excited about this paper. Excited to hear people’s thoughts. Link: institute.deepmind.com/essay…
152
47
26
468
35,010
Captain Picard had an ineffable influence on a young me. There's so many good moments from 'The Drumhead' alone but this scene is seared into my brain. I remember the first time I saw it and how it would come to form my moral compass. Rest In Peace Patrick Stewart.
44
424
37
3,151
82,596
Frontier models are much better than human reviewers. I trust the methodology below far more than any journal review process. saez-zucman-audit-replicatio…
9
7
4
82
11,804
Academics have to invest tremendous effort into a single paper, so we often work on papers that we suspect will yield interesting & significant results. But that also pushes us to predictable work. AI has no such constraints, forcing us to rethink a lot of how science is done.
New working paper: Can AI forecast treatment effects in economics from pre-registration documents? Answer: Yes. 1/
182
42
26
420
37,623
"I pay attention to NOTHING" 🤣 Dundee boss Steven Pressley's view on social media... #BBCFootball
29
42
27
681
112,517
Absolutely stunning essay by @DouglasKMurray about the crisis in France. It includes this quote. thefp.com/p/douglas-murray-f…
130
1,348
83
7,318
287,974