From storyflo. This is your daily audio brief. Theo here. August 11th, tech desk. Five stories from the last twenty-four hours — here's where I'd start. Let's get into it. First, from KDnuggets. 3 Visual Proofs of the Central Limit Theorem to Build Your Intuition. To build your intuition, this article shows three visual proofs that the classic bell curve appears in myriad situations. The central limit theorem (CLT), in very broad terms, tells us that the "magical" bell curve of normal distributions happens (and it does, a lot!) in the real world, often regardless of the data's original shape. But have you wondered why? The bottom line behind CLT is a fundamental statistical rule: if you take enough samples from any data and calculate their averages, these averages will approach a normal distribution — no matter what the original data's form was. Next. Second, from KDnuggets. The Ultimate Guide to Contributing to Open Source Projects. GitHub added 36 million new developers in 2025, roug

From storyflo. This is your daily audio brief. Theo here. August 11th, tech desk. Five stories from the last twenty-four hours — here's where I'd start. Let's get into it. First, from Latent.Space. [AINews] Muse Glimmer and Spark: Open Weights return Personal Superintelligence promise. Last week was the 1 year anniversary of Zuck’s original Personal Superintelligence essay, and MSL seems to be feeling a second wind this year, as they slowly ramped up with the Dreamer acquisition and then Muse Spark and recently Muse Code. For a while it seemed like MSL was being rather timid with the launches… but today that all changed. Zuck returned with a hit sequel essay and released MSL’s first real open weights frontier-ish small LLM, with Spark to also be released soon. Next. Second, from Agentic AI. Black Hat and DEF CON 34 Trip Report: What Vegas Actually Shipped in Agentic AI Security. I spent the first week of August in Las Vegas. Black Hat USA ran August 1 through 6 at Mandalay Bay, DEF CON

From storyflo. This is your daily audio brief. Hey, it's Theo. August 10th. Here are five stories I'd flag if you missed yesterday's end-of-day. Let's get into it. First, from KDnuggets. Top 10 AI Influencers of 2026. Ten AI influencers who are actually shaping 2026, from safe superintelligence to AI-native search. Here is who to follow and why. AI is no longer a novelty; it's the foundational layer of modern computing. As the technology matures, the voices guiding its development — from frontier model builders to hardware architects and safety advocates — matter more than ever. Next. Second, from KDnuggets. Specification Engineering: The New Skill After Prompt Engineering. Prompt engineering taught us how to ask better questions. Specification engineering teaches us how to define better work. For the last two years, people have learned to get better results from large language models (LLMs) by writing clearer prompts: add context, define the role, give examples, specify the format, an

From storyflo. This is your daily audio brief. Hey, it's Theo. August 10th. Here are five stories I'd flag if you missed yesterday's end-of-day. Let's get into it. First, from Cryptonite Weekly Rap. The New Gov AI OS & Agents. AI as a company (and public agency) OS To adapt to rapid AI efficiency gains, Jack Dorsey cut 40% of his workforce at Cash App’s parent company, Block, despite a strong 2025. Rather than engage in repeated rounds of cuts that invariably damage morale, Jack decided to embrace the new AI reality all at once. “AI doesn’t just automate tasks—it rewrites the e… Next. Second, from InfoQ. Beyond Consensus: The Fragmentation of AI Policy Across the Linux Ecosystem. I’ve been chewing on how the Linux world is handling AI, and it turns out there isn’t a single playbook. The GNU Compiler Collection keeps things tight, refusing to let AI‑generated code slip in without a lot of checks, while the kernel itself is more about getting things done—its policy is pragmatic, letting

From storyflo. This is your daily audio brief. Theo, August 9th. The systems update — five tech stories that bear on what's coming next. Let's get into it. First, from The Decoder. Google dismantles Deepmind and bets on a fresh start as Hassabis heads for the exit. Google is pulling DeepMind back into the core of Alphabet, stripping its independent status and shifting Gemini work to the Bay Area. Demis Hassabis is expected to leave in the next few months, while senior researcher Koray Kavukcuoglu will run day‑to‑day ops without a formal CEO title. Internally the move reflects deeper friction: teams are hitting snags training the newest frontier models, even though Google Cloud is still pulling in billions. The real question now is whether the company is betting on tighter infrastructure control or simply lagging behind the fast‑moving AI leaders. Next. Second, from The Decoder. AI's energy appetite drives Nvidia and Amazon to pour billions into massive power infrastructure. The AI indu

From storyflo. This is your daily audio brief. Theo, August 9th. The systems update — five tech stories that bear on what's coming next. Let's get into it. First, from Hackaday. Tracking Bees with a Stereo Camera. For whatever reason, [Jeremy] wants to photograph bees in flight. This is one of those things that doesn’t seem that hard until you try to do it. He’s got the mechanical part that can move the camera as fast as the bee flies. But you still have to point the camera correctly. He considered quite a few methods of acquiring and tracking the bee, but he finally settled on a stereo camera. His requirements also made the camera design challenging. In particular, the bees move fast enough that things like USB transmission times become significant. You can see his solution in the video below. Next. Second, from XDA Developers. Microsoft accidentally rolled out a OneDrive app to everyone's PCs, and there's no simple way to fix it. As big as Microsoft is, it can make mistakes sometimes

From storyflo. This is your daily audio brief. Theo, August 8th. The systems update — five tech stories that bear on what's coming next. Let's get into it. First, from The Decoder. AI agents use roughly 600 times more energy than a simple chat prompt. So I was reading about this climate scientist, Zeke Hausfather, who tracked his usage of this AI model called Claude Code over eight weeks. He found out that for every prompt he gave it, it used about 600 times more energy than a typical AI chat. Now, Google and OpenAI have been reporting pretty low energy figures for their AI agents, but Zeke's data shows that's not even close to the truth. He used 3.2 billion tokens and about 170 kilowatt-hours of electricity from the data center. It's crazy to think about how much energy is actually being used behind the scenes when we interact with these AI agents. Next. Second, from The Decoder. Fields Medalist who published a paper on AI-driven human extinction now works for OpenAI. So I was reading

From storyflo. This is your daily audio brief. Theo, August 8th. The systems update — five tech stories that bear on what's coming next. Let's get into it. First, from Emerging AI. Master the Claude 5 Family. Claude 5 is powerful enough to save you a week of work. It is also powerful enough to spend hours carrying the wrong instructions, using unnecessary tools, and finishing a job you did not actually ask for. That is the first thing I learned from working with this new family. Fable 5 can remain inside a difficult project for hours. Opus 5 can inspect a codebase, make architectural decisions, build across many files, run tools, review its own work, and correct mistakes. Next. Second, from mtslive. 8/7: OpenAI Delays Astra. It’s Friday. Between OpenAI disproving 10 novel math conjectures, Qwen3.8, the White House AI framework, FLUX 3 Video, the UK AISI report on cyber misalignment, Meta Muse Spark 1.2, Anthropic’s in-house chip team, Google DeepMind’s massive reorg, Terafab’s groundbr

From storyflo. This is your daily audio brief. Theo, August 7th. The systems update — five tech stories that bear on what's coming next. Let's get into it. First, from KDnuggets. 5 Free Courses to Learn Modern AI and LLMs. Learn how to use generative AI at work, build RAG and agentic apps, fine-tune models, work with the Hugging Face ecosystem, and prototype AI products with hands-on resources. # The Path to Learning AI & LLMs There are a lot of AI courses online, but not all of them teach modern AI. Some courses have only changed their title from "machine learning" to "AI," while the content is still mostly the same. But modern AI is different. Next. Second, from KDnuggets. Small Language Models with Hugging Face transformers Library + smolLM3. Running a 70B model in production is expensive, and for many tasks, unnecessary. If you're building a focused pipeline, a well-trained 3B model will match or beat the 70B on your specific task at a fraction of the cost. Up next. Third, from AI

From storyflo. This is your daily audio brief. Theo, August 7th. The systems update — five tech stories that bear on what's coming next. Let's get into it. First, from Latent.Space. [AINews] AMD buys Taalas. In The Custom ASIC Thesis we said Taalas was worth paying attention to, and in the Inference Inflection we said everything would go vertical. Our Baseten episode had some skeptical counterpoints against etched LLMs, not just custom ASICs, but clearly Lisa Su disagrees for now. Congrats! AI News for 8/5/2026-8/6/2026. We checked 12 subreddits, 544 Twitters and no further Discords. AINews’ website lets you search all past issues. As a reminder, AINews is now a section of Latent Space. Next. Second, from Mashable Tech. Mashable will send the best deals directly to your phone daily — how to sign up for free. They're constantly scanning top retailers like Amazon, Walmart, and Best Buy to find limited-time discounts on products they'd actually recommend. You can expect to get a Deal of t