Shawn Sullivan (@shawntsullivan) header
Shawn Sullivan (@shawntsullivan) profile picture

CTO @ https://xitter.gq/t.co/Ejx8na6sMZ: GenAI for Edu. Benchmarking AI in Edu. Early reading with Reading Critters. Co-founder/ex-CTO @ Phase Genomics: AI+Genomics. MIT CS, Ex-MSFT

Seattle
Joined May 2014
Photo by Shawn Sullivan (@shawntsullivan)Photo by Shawn Sullivan (@shawntsullivan)Photo by Shawn Sullivan (@shawntsullivan)Photo by Shawn Sullivan (@shawntsullivan)Photo by Shawn Sullivan (@shawntsullivan)Photo by Shawn Sullivan (@shawntsullivan)Photo by Shawn Sullivan (@shawntsullivan)Photo by Shawn Sullivan (@shawntsullivan)Photo by Shawn Sullivan (@shawntsullivan)Photo by Shawn Sullivan (@shawntsullivan)
MLB is cooked If your baseball understanding is a spreadsheet, @TheJudge44 is your guy If you actually watch baseball, you know @Mariners #CalRaleigh turned in a season no catcher has ever come close to Judge was great. Cal was irreplaceable. "Valuable" is not "cell A1"
46
Some actual data-driven, practical advice for students using AI to get ready for AP test season, including specific recommendations for which models to use for which subjects. Edu AI benchmarks that mean something!
AP exam season is looming! 😰 Feeling the pressure? We tested how well popular AI models can help you get a 5. The results might surprise you 🧵👇 edubench.com/edublog/which-a… #APTests #AI #EduTech 1/8
2
142
Tbh I was surprised to see these results, given most LLMs did reasonably well creating multiple-choice questions. Something about creating a quiz really seems to throw them for a loop... digging into why soon.
Now live on edubench.com: Quiz Composition benchmarks! Unlike MCQ Generation, frontier models struggle with most aspects of creating good quizzes. More details soon; for now, head over to edubench.com and see for yourself. What stands out to you?
93
AI-generated MCQs can look good but still fail students. A key issue? Bad distractors—wrong answers that are too obvious to provide effective assessment. Measuring distractor quality is a critical tool for using AI effectively in the classroom
We're benchmarking AI for education. What does that mean? Let's dive into an example. Suppose you're an AP World History teacher who wants to use an LLM to create a quiz question for your students... 🧵 1/7
70
One takeaway from our recent comparison of Claude 3.5 and 3.7 for edu: Both Claude versions fail because they're optimized for capability demos, not educational outcomes. Simply being able to generate questions isn't enough. We need models that create content that actually helps students learn. #EdTech
We analyzed Claude 3.5 vs 3.7 for creating educational content. The results are surprising... edubench.com/edublog/claude-… 1/
1
100
Excited to announce a new project! In EduBench, we aim to comprehensively benchmark AI in educational settings. We're starting with multiple-choice questions. Much more to come, including our own efforts to create best-in-class LLMs for edu applications. Follow for more!
🚨 AI in education is at a crossroads. Too many AI tools look helpful but actually mislead students, fail to align with curricula, and give a false sense of competence. We’re launching EduBench to demand real educational impact from AI. 🧵👇
75
Holy sh*t Been using @lovable for 2 weeks and it's already a game changer. Looks like it's only going to keep getting better
Bringing pixel perfect editing to your @lovable apps (feature preview)
1
155
Shawn Sullivan retweeted
NEW POST My colleague Bharani Subramaniam has started to write patterns from our recent work building production Gen AI applications. We begin with Evals - ways of assessing if they are working effectively. martinfowler.com/articles/ge…
2
36
1
175
19,571
Strong agreement: Ed Dept's AI in education report hits the nail on the head. AI isn't just a cost-cutting tool; it's the key to unlocking personalized learning at scale. We can finally tailor education to each child's needs, not just the average student. tech.ed.gov/ai-future-of-tea…
69
Want to future-proof education? Start by involving educators in AI development. They're not just end-users; they're the experts we need to build truly effective educational AI. tech.ed.gov/ai-future-of-tea…
1
64
The biggest risk in educational AI isn't adopting it—it's adopting it without a clear, shared vision of what education should be. Tech serves goals, not the other way around. Strong agree with this point in Ed report tech.ed.gov/ai-future-of-tea…
2
47
Reducing bias in education isn't just about human awareness anymore. AI-powered assessment tools, when properly designed, could be our best shot at truly fair evaluation.
38
AI in education isn't a silver bullet, but it's close. It can provide the personalized, adaptive learning that's been the holy grail of education for decades. The potential to close achievement gaps and unleash every child's potential is too great to ignore.
46
Proactive policy on AI in education isn't just smart - it's necessary. We're on the cusp of an educational revolution. Those who embrace AI responsibly will lead; those who don't will be left behind. The future of our children's education is at stake tech.ed.gov/ai-future-of-tea…
38
Here's a radical thought: What if AI could make teaching one of the most sought-after professions? By handling the grunt work, AI lets teachers focus on what they do best—inspire and guide.
1
2
265
This DoE report talks about 'human in the loop' AI. We believe in 'child in the spotlight' learning – where technology adapts to each learner's unique needs, interests, and pace. insidehighered.com/news/quic…
75
Some really interesting concepts here. Most are fractal too. You can apply them at any scale. Good reading anyone can apply to their own work and life
The way that Jensen Huang runs Nvidia is wild: 40 direct reports. No 1:1s. No formal planning cycles. And no status reports. In a recent interview, he went in-depth on his Leadership style. Every entrepreneur must understand why it works:
113
Want your kids to be competitive in tomorrow's job market? Make sure they're learning alongside AI, not just about it. Practical experience with AI tools is becoming as crucial as basic math or writing skills.
40
The 3 Rs of education are outdated. Today's students need Reading, 'Riting, 'Rithmetic, and AI. Career readiness in the 21st century means understanding how to leverage AI tools responsibly.
41
DoE report misses a crucial point: AI in education isn't just about content delivery. It's about creating emotional connections that motivate children to love learning. As Asimov predicted, children who grow up with AI will form emotional attachments to it insidehighered.com/news/quic…
1
87