Nicole Hennig
LightNews LightNews
banner
nic221.bsky.social
Nicole Hennig
@nic221.bsky.social
4.9K followers 400 following 3.8K posts
E-learning dev & AI educator at U of Arizona Libraries. Former head of UX at MIT Libraries. Winner of MIT Excellence Award. nicolehennig.com. Digital nomad from 2013-17, locationflexiblelife.com.
- vegetarian
- car-free
- universal basic income: yes
Posts Replies Media Videos
Pinned
nic221.bsky.social
Nicole Hennig @nic221.bsky.social · Nov 13
Generative AI News https://nicolehennig.com/gen-ai-news/ (my free monthly newsletter) #AI #Substack #newsletters
10 Tips for Keeping Up with Generative AI
nic221.bsky.social
Paper Notes: Nested Learning itcanthink.substack.com/p/paper-notes-… #AI #research #Google
As usual though, with a theory paper like this, it's worth noting that these seem fairly far from use on anything more than toy tasks — and that inference with transformers continues to improve, in part because they scale extremely well. Titans and HOPE won't be replacing GPT next year or anything. But the idea here — that we're thinking about neural networks wrong, that "optimizers" and "modules" are not actually discrete entities — seems extremely interesting and has a ton of potential for the future.
November 26, 2025 at 6:46 AM
nic221.bsky.social
Six reasons to think there’s an AI bubble — and six reasons not to www.understandingai.org/p/six-reasons-… #AI #bubble #boom
Last week, we were discussing how one could capture the entire Al-bubble debate in about 12 statistics that people just keep citing and reciting - on CNBC, on tech podcasts, in Goldman Sachs Research documents, and at San Francisco AI parties. Since everybody seems to be reading and quoting from the same skinny playbook, we thought: What the hell, let's just publish the whole playbook! If you read this article, we think you'll be prepared for just about every conversation about Al, whether you find yourself at a Bay Area gathering with accelerationists or a Thanksgiving debate with Luddite cousins. We think some of these arguments are compelling. We think others are less persuasive. So, throughout the article, we'll explain both why each argument belongs in the discussion and why some arguments don't prove as much as they claim. Read to the end 1, and you'll see where each of us comes down on the debate.
November 26, 2025 at 4:10 AM
nic221.bsky.social
anthropics-claude-opus-4-5-is-here-cheaper-ai-infinite-chats-and-coding venturebeat.com/ai/anthropics-… #AI #Claude #coding
November 26, 2025 at 2:27 AM
nic221.bsky.social
How GPT-5 helped mathematician Ernest Ryu solve a 40-year-old open problem openai.com/index/gpt-5-ma… #AI #math #GPT5
Text Shot: It was a surreal feeling,” he said, equating the experience to working with a competent student who would come up with ideas, ask questions, and brainstorm. “It was really intellectually exciting to work with GPT‑5,” Ryu said. “I had to be vigilant in checking the calculation steps, but it was still much faster than doing the calculation myself.” While the process was significantly faster, it also required careful verification. GPT‑5 often produced arguments that looked plausible but did not hold up on inspection. Ryu reviewed each suggestion, discarded the ones that failed, and developed the ones that contained a mathematically meaningful idea. The workflow depended on his judgment—both to identify promising directions and to recognize when a line of reasoning had been fully explored. Ryu used GPT‑5 like a tool, and he took great care to understand what it was doing so he could discern what it was good at and what it wasn’t. When asking GPT‑5 to check the work, Ryu found…
November 26, 2025 at 2:24 AM
nic221.bsky.social
Introducing shopping research in ChatGPT openai.com/index/chatgpt-… #AI #ChatGPT #shopping
Text Shot: Shopping research is built for that deeper kind of decision-making. It turns product discovery into a conversation: asking smart questions to understand what you care about, pulling accurate, up-to-date details from high-quality sources, and bringing options back to you to refine the results. It performs especially well in detail-heavy categories like electronics, beauty, home and garden, kitchen and appliances, and sports and outdoor.
November 25, 2025 at 11:49 PM
nic221.bsky.social
googles-nested-learning-paradigm-could-solve-ais-memory-and-continual venturebeat.com/ai/googles-nes… #AI #memory #Google
Text Shot: The problem is that today’s transformer-based LLMs have no mechanism for “online” consolidation. Information in the context window never updates the model’s long-term parameters — the weights stored in its feed-forward layers. As a result, the model can’t permanently acquire new knowledge or skills from interactions; anything it learns disappears as soon as the context window rolls over. A nested approach to learning Nested Learning (NL) is designed to allow computational models to learn from data using different levels of abstraction and time-scales, much like the brain. It treats a single machine learning model not as one continuous process, but as a system of interconnected learning problems that are optimized simultaneously at different speeds. This is a departure from the classic view, which treats a model's architecture and its optimization algorithm as two separate components.
November 25, 2025 at 11:35 PM
nic221.bsky.social
Suno, Yout, Perplexity AI and §1201: AI Training and another piece of the DMCA authorsalliance.substack.com/p/suno-yout-pe… #AI #Suno #copyright (another excellent piece from Authors Alliance)
Continued AI development shouldn't hinge on a misapplied anti-circumvention rule. If copyright owners want to challenge Al training, they should do so directly on copyright infringement grounds or through Congress. Not through a legal side door designed to regulate encryption in 1998.
November 24, 2025 at 2:55 PM
nic221.bsky.social
ElevenLabs — Introducing ElevenLabs Image & Video elevenlabs.io/blog/introduci… #AI #video
Text Shot: Within ElevenLabs, you can now bring ideas to life in one complete creative workflow. Use leading models like Veo, Sora, Kling, Wan and Seedance to create high-quality visuals, then bring them to life with the best voices, music, and sound effects from ElevenLabs.
November 24, 2025 at 5:55 AM
nic221.bsky.social
Generative UI: A rich, custom, visual interactive user experience for any prompt research.google/blog/generativ… #AI #Google #Gemini
Text Shot: We introduce a novel implementation of generative UI, enabling AI models to create immersive experiences and interactive tools and simulations, all generated completely on the fly for any prompt. This is now rolling out in the Gemini app and Google Search, starting with AI Mode.
November 24, 2025 at 2:29 AM
nic221.bsky.social
Hyperproductivity: The Next Stage of AI? secondthoughts.ai/p/hyperproduct… #AI #coding #workplace #optimizing
1. They are aggressively using AI to accelerate their work. 2. They aren't just using off-the-shelf tools like ChatGPT or Claude Code. They're building bespoke productivity tools customized for their personal workflows - and, of course, they're using AI to do it. 3. Their focus has shifted from doing their jobs to optimizing their jobs. Each week, instead of delivering a new unit of work, they deliver a new improvement in productivity. 4. Their work builds on itself: they use their Al tools to improve their AI tools, and the work they're optimizing includes the optimization process.
November 23, 2025 at 3:12 PM
nic221.bsky.social
Microfluidics: The Secret to Cooler, Faster AI Chips spectrum.ieee.org/microfluidics-… #AI #DataCenters #cooling #advancements
Text Shot: Corintis, a Swiss company, is developing a technology called microfluidics, in which water or another cooling liquid is channeled directly to specific parts of a chip to prevent overheating. In a recent test with Microsoft, servers running the company’s Teams video conferencing software recorded heat removal rates three times as efficient as other existing cooling methods.Compared to traditional air cooling, microfluidics lowered chip temperatures by more than 80 percent.
November 23, 2025 at 2:48 PM
nic221.bsky.social
Will there be a billion nomads by 2035? globalnatives.substack.com/p/billion-noma… #DigitalNomads #borders #RemoteWork (this is interesting to me, especially since I lived as a digital nomad from 2013-2018)
When the first wave of nomads set out, they weren't trying to start a movement. They just wanted freedom-from commutes, office politics, and unpleasant weather. But in choosing mobility, they revealed something bigger: that work didn't have to be tied to a single location. That idea, once radical, is becoming increasingly normalised in global culture. It's still spreading-unevenly, but irreversibly-as people who never imagined a nomadic life realise that location can be a personal choice based on individual priorities. So will Levels' prediction come true? Will there really be a billion nomads by 2035? Yes—but not in the way people imagined when he said it. The nomad movement is no longer an alternative to the mainstream, but a signal of where the mainstream is heading. This
November 23, 2025 at 2:15 PM
nic221.bsky.social
AI Blackmail: Fact-Checking a Misleading Narrative www.aipanic.news/p/ai-blackmail… #AI #Anthropic #journalism
AI Blackmail: Fact-Checking a Misleading Narrative
CBS should issue a correction to its “60 Minutes” show on Anthropic.
www.aipanic.news
November 23, 2025 at 4:24 AM
nic221.bsky.social
Why it takes months to tell if new AI models are good www.seangoedecke.com/are-new-models… #AI #evals #benchmarks #vibes
November 23, 2025 at 2:46 AM
nic221.bsky.social
Nano Banana comics for Caulfield’s “Three moves, seven tips” - Nicole Hennig nicolehennig.com/nano-banana-co… #AI #NanoBananaPro
Text Shot: Now this post is not about how to use his method… instead it’s for showing my fun experiments with using Nano Banana Pro to illustrate his ideas. Since Google just released Nano Banana Pro recently, I’ve been experimenting with using it to create infographics and comic panels. It works well! I use it in the paid version of Gemini. So I thought I’d use the “three moves, seven tips” idea to experiment with. Here are some of my results.
November 22, 2025 at 11:28 PM
nic221.bsky.social
The False Hope of Content Licensing at Internet Scale - Matthew Sag www.promarket.org/2025/11/19/the… #AI #copyright
Text Shot: The impossibility of clearing rights at scale does not settle the normative and doctrinal arguments for and against treating AI training as fair use, but I have said enough about why nonexpressive use should be fair use elsewhere. This article addresses just one small piece of the debate, whether a private ordering solution is a feasible alternative to fair use. Despite the most ardent wishful thinking of copyright maximalists and private ordering enthusiasts, it seems quite unlikely that cutting-edge LLMs could be trained on a mixture of public domain and licensed materials without access to the volumes of training data developers have scrapped from the open Internet.
November 22, 2025 at 3:00 PM
nic221.bsky.social
Nano Banana Pro: But Did You Catch These 10 Details? www.youtube.com/watch?v=9hv4nr… #AI #images
November 22, 2025 at 3:43 AM
nic221.bsky.social
SIFT for AI: Introduction and Pedagogy - Mike Caulfield mikecaulfield.substack.com/p/sift-for-ai-… #AI #InformationLiteracy #education
I think a possible reaction to this is that the students are not thinking at all. After all, isn't it the LLM that is coming up with these alternative interpretations and these charts of evidence? Seems lazy, right? But I think this gets it backwards. There is no way that any student without significant knowledge is going to be able to surface these possibilities. The ability to look at a snippet of film like this and surface three or four perfect lenses on it, or to look at a research paper on chocolate and surface the six most important questions about the methods-findings fit is a level of expertise that people gain only after years of experience. But once these things are surfaced students are capable of engaging in verification, evaluation, synthesis and exploration. And exposed to that over time with guidance and a decent set of educational materials, students will start to build the foundational knowledge, skills, and understandings to do it on their
November 22, 2025 at 2:55 AM
nic221.bsky.social
HypeDetector: # Linguistic Analysis: Misleading Language and Rhetorical Techniques Based on my analysis of the TechCrunch article "Rising data center electricity use risks blackouts during winter sto poe.com/s/HUi70cMw69tE… (my bot for detecting misleading language) #AI #hype #NewsCoverage
HypeDetector By @nic221 Give it a news story and it detects misleading language or rhetorical techniques.
November 21, 2025 at 3:21 PM
nic221.bsky.social
WMG settles Udio lawsuit, strikes licensing deal for ‘next-generation’ AI music platform coming in 2026 - Music Business Worldwide www.musicbusinessworldwide.com/warner-music-s… #AI #music #copyright #settlements
Text Shot: The companies have also entered into a licensing deal for a “next-generation” AI–powered music creation, listening, and discovery platform set to launch next year. The news arrived just an hour after WMG announced a new partnership with Stability AI on Wednesday (November 19), which the companies say will “advance the use of responsible AI in music creation”. According to a press release, the agreement with Udio spans WMG’s recorded music and music publishing businesses and “creates new revenue streams for artists and songwriters, while ensuring their work remains protected”. The settlement and licensing deals come just over a year after the RIAA, on behalf of all three major record companies, sued Udio and rival AI platform Suno for “mass infringement” of copyright. Universal Music Group also settled with Udio last month, in addition to signing a deal for a licensed AI music platform set to launch in 2026.
November 21, 2025 at 3:07 PM
nic221.bsky.social
Scholar Labs Early Review: Google Scholar Finally Enters the AI Era - Aaron Tay aarontay.substack.com/p/scholar-labs… #AI #GoogleScholar
Scholar Labs Early Review: Google Scholar Finally Enters the AI Era
Generated by Nano-Banana Pro from text of this blog post
aarontay.substack.com
November 21, 2025 at 5:57 AM
nic221.bsky.social
Google Has Your Data. Gemini Barely Uses It. shloked.com/writing/gemini… #AI #context #memory #Gemini
Text Shot: If context is the new battleground, Google starts with a structural advantage no one else has: your mail, docs, calendar, photos, maps, browser, even your phone. ChatGPT and Claude are still begging for slivers of that graph through connectors; Google already sits on the whole thing. And to their credit, the pipes are technically there. With a single toggle in settings, you can let Gemini see Gmail, Drive, Docs, and the rest without messy OAuth flows or permission mazes. Once connected, it can already do useful things: summarize your inbox, draft calendar invites, pull in details from a document. The Google Workspace connector hides in settings The problem is that Gemini doesn't really lean on any of this. ChatGPT's Pulse keeps nudging you to connect your calendar. Claude repeatedly insists you "bring your docs in here and chat with them." Gemini, by contrast, hides the workspace connector in settings and then tiptoes around it. The ability to fuse your life into the…
November 21, 2025 at 4:28 AM
nic221.bsky.social
Fix the News 320: Aha! Lost Worlds. Mongolia Colombia. Tuberculosis. Uninterrupted toolmaking. fixthenews.com/p/320-aha-lost… #news #journalism
Of course, this has been going on for a long time. William Randolph Hearst coined the phrase "if it bleeds, it leads" back in the 1890s, during the days of yellow journalism. What's different now is that we all get our news on the black devil glass, all the time, as it happens, from everywhere - which is why we are drowning in an endless stream of alarm that makes the world feel far more dangerous than it actually is. We've said this a hundred times before, but it's always worth repeating: the world's biggest news organisations aren't reflecting the world. They are reflecting what's dramatic, rare, and cheap enough to turn into a story. If you're after an accurate picture of what's getting worse, or what's improving, you'll need to look somewhere other than the news.
November 21, 2025 at 3:49 AM
nic221.bsky.social
Q3 Global Power Report: No fossil fuel growth expected in 2025 | Ember ember-energy.org/latest-insight… #energy #renewables
Text Shot: No growth in fossil power is expected in 2025 as clean power growth meets all new demand. Record growth in solar power, combined with moderate wind growth, exceeded the rise in electricity demand in the first three quarters of 2025, even when accounting for a fall in hydropower. As a result of fast-rising clean power, fossil generation in the global power sector is expected to remain flat in 2025 for the first time since the Covid-19 pandemic.
November 21, 2025 at 3:44 AM
nic221.bsky.social
7 tips to get the most out of Nano Banana Pro blog.google/products/gemin… #AI #Google #NanoBanana #images
Text Shot: We’ve given Nano Banana a massive upgrade. Built on Gemini 3, Nano Banana Pro is our most advanced image model to date, ready to bridge the gap between imagination and professional execution. Available now in the Gemini app and starting to roll out in AI Studio, Vertex and more, it features state-of-the-art text rendering in over multiple languages, and advanced controls like the ability to input up to 14 images into a composition (varies by surface).
November 20, 2025 at 11:59 PM