My opinions only here. 👨‍🔬 RS DeepMind Past: 👨‍🔬 R Midjourney 1y 🧑‍🎓 DPhil AIMS Uni of Oxford 4.5y 🧙‍♂️ RE DeepMind 1y 📺 SWE Google 3y 🎓 TUM 👤 @nwspkhttps://bsky.app/profile/blackhc.bsky.socialhttps://bsky.app/profile/blackhc.bsky.social/post/3mhd2hi5eio2uA while back, Andrej Karpathy said the app store will be replaced by generated, disposable software," and Amjad Masad predicted that the value of all application software will go to zero I think this "ephemeral software hypothesis" is wrong, though, and I want to explain why:18 Mar 2026 08:19 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3mhd2hi5eio2uhttps://bsky.app/profile/blackhc.bsky.social/post/3lrdu5gmefs2uWhat's your favorite Veo video?11 Jun 2025 16:44 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lrdu5gmefs2uhttps://bsky.app/profile/blackhc.bsky.social/post/3lr7eafp5tz22I'm late to review the "Illusion of Thinking" paper, so let me collect some of the best threads by and critical takes by @scaling01 in one place and sprinkle some of my own thoughts in as well. The paper is rather critical of reasoning LLMs (LRMs): https://x.com/MFarajtabar/status/193070759164849373009 Jun 2025 21:49 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lr7eafp5tz22https://bsky.app/profile/blackhc.bsky.social/post/3lpehv5xyzi2lI want to share my latest (very short) blog post: "Active Learning vs. Data Filtering: Selection vs. Rejection." What is the fundamental difference between active learning and data filtering? Well, obviously, the difference is that: 1/1117 May 2025 11:47 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lpehv5xyzi2lhttps://bsky.app/profile/blackhc.bsky.social/post/3lni36i3jy22iI want to share a blog post on our paper "All Models are Wrong, Some are Useful: Model Selection with Limited Labels" which we will present at AISTATS 2025 next week With @pokanovic.bsky.social‬, Jannes Kasper, @thoefler.bsky.social, @arkrause.bsky.social, and @nmervegurel.bsky.social23 Apr 2025 11:20 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lni36i3jy22ihttps://bsky.app/profile/blackhc.bsky.social/post/3lm2vbtvhb32qI want to reshare @brandfonbrener.bsky.social's @NeurIPSConf 2024 paper on CoLoR-Filter: A simple yet powerful method for selecting high-quality data for language model pre-training! With @hlzhang109.bsky.social @schwarzjn.bsky.social @shamkakade.bsky.social05 Apr 2025 12:04 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lm2vbtvhb32qhttps://bsky.app/profile/blackhc.bsky.social/post/3lm2vbtvhb32qI want to reshare @brandfonbrener.bsky.social's @NeurIPSConf 2024 paper on CoLoR-Filter: A simple yet powerful method for selecting high-quality data for language model pre-training! With @hlzhang109.bsky.social @schwarzjn.bsky.social @shamkakade.bsky.social05 Apr 2025 12:04 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lm2vbtvhb32qhttps://bsky.app/profile/blackhc.bsky.social/post/3lf6ovnuwpk23Ever wondered why presenting more facts can sometimes *worsen* disagreements, even among rational people? 🤔 It turns out, Bayesian reasoning has some surprising answers - no cognitive biases needed! Let's explore this fascinating paradox quickly ☺️07 Jan 2025 22:25 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lf6ovnuwpk23https://bsky.app/profile/blackhc.bsky.social/post/3lf6ovnuwpk23Ever wondered why presenting more facts can sometimes *worsen* disagreements, even among rational people? 🤔 It turns out, Bayesian reasoning has some surprising answers - no cognitive biases needed! Let's explore this fascinating paradox quickly ☺️07 Jan 2025 22:25 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lf6ovnuwpk23https://bsky.app/profile/blackhc.bsky.social/post/3lerkfxxl4n2lI didn't talk about it but I also made heavy use of Claude 3.5 and also o1 and Gemini when creating my lecture series on info theory and active learning in 3.5 weeks: https://bsky.app/profile/blackhc.bsky.social/post/3ldiav6g6vc2602 Jan 2025 17:00 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lerkfxxl4n2lhttps://bsky.app/profile/blackhc.bsky.social/post/3ldiav6g6vc26The slides for my lectures on (Bayesian) Active Learning, Information Theory, and Uncertainty are online now 🥳 They cover quite a bit from basic information theory to some recent papers: blackhc.github.io/balitu/ and I'll try to add proper course notes over time 🤗17 Dec 2024 06:50 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3ldiav6g6vc26https://bsky.app/profile/blackhc.bsky.social/post/3lddsybmg3s2kThanks for following me here! 🫶 I went through my notifications to follow people if they are in ML research, doing PhDs, etc, to have a nice feed focused on ML. Apologies to anyone I have missed! You can unfollow and refollow me to give me a new notification (I suppose)! Plz update your profiles 🙏15 Dec 2024 12:30 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3lddsybmg3s2khttps://bsky.app/profile/blackhc.bsky.social/post/3ld6qig6p5k2kThanks for the support! I hope I am doing the right thing here. But having written pseudo-code for both approaches now, it does seem like the only addition is a `.flatten()` 🙃 https://openreview.net/forum?id=0NMzBwqaAJ [contains quote post or other embedded content]13 Dec 2024 12:03 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3ld6qig6p5k2khttps://bsky.app/profile/blackhc.bsky.social/post/3ld3yjj7a3s24Visualization of epistmic uncertainty via MI as the gap between average conditional entropy and predictive entropy: https://blackhc.github.io/balitu/lectures/lecture_3_bnns.html#/interactive-plot12 Dec 2024 09:48 +0000at://did:plc:sux7xhmscgoe3cd4c2oe7aec/app.bsky.feed.post/3ld3yjj7a3s24