Weekly Link Roundup #55
I can't drive........... (IFYKYK)
Not a Coder? With A.I., Just Having an Idea Can Be Enough: (gift link to story) Let me just say that I’m including this article for two reasons. The first is that I think this is an important dynamic - allowing people who aren’t software developers to be able to create apps. I think its a dynamic that has pros and cons. Pros = more people creating and less translation loss. Cons = more people creating not just apps but potential security holes, potential tech debt, a graveyard of orphaned apps. The second reason I include this article is that I absolutely HATE the term “vibecoding.” The article says it’s “useful shorthand for the way that today’s A.I. tools allow even nontechnical hobbyists to build fully functioning apps and websites, just by typing prompts into a text box.” That’s BS. The term is a gate keeper. We don’t have a parallel term for when coders sit around and try their darndest to come up with a non-coding idea…maybe “vibeideaing” just to show that its not really all that valuable. Stop it.
Estonia launches AI in high schools with US tech groups: This is smart, upstream thinking > > “Estonia is teaming up with OpenAI and Anthropic to launch a nationwide drive to teach artificial intelligence skills to high school students, aiming to help prepare them for jobs of the future. The initiative, known as AI Leap, draws on the Baltic country’s public digital infrastructure built over the past 30 years and its strong educational culture. Estonia is ranked top among European countries in the international Pisa education tests. Estonia’s President Alar Karis said the initiative was not intended to replace teachers in the classroom but to develop critical thinking among students and awareness of AI.”
xAPI was designed to break down silos. So, don’t put it in a silo: “So, here comes xAPI and the ability to emit commonly structured data from any source — including the LMS, the mobile app, and the sim (and whatever else came along). And it seems great — in theory. But, the learning curve is real. And any sort of data integration — let alone semantic data integration dependent upon a bulletproof identity management protocol — is notoriously difficult. And when those integrations don’t come together the way that you think they will, the result is a lot of bad xAPI data.”
Sweden’s Lovable, an app-building AI platform, rakes in $15M after spectacular growth: OK...here's a free idea for #LearningAndDevelopment teams...pick a low/code AI-enabled, app-building platform like Loveable. Now go out to each of the business units (technical and non-technical) and get a clear picture of their needs/requirements and so on...something that as a Learning team, you should probably be doing anyway. Now, go to the CIO and get clearance to use these tools in-house and get clear on the parameters - e.g. which data can be touched etc. Now coming the business unit requirements with some #training content that teaches the app-building platform. Now launch your innovation contest in which people don't just submit ideas but actual prototype or working apps. Now your Learning team is really connected to the business, you've shown how innovative you are and the value from those apps that actually get deployed, leads right back to you. > > "Lovable enables anyone to build what it calls production-ready software without needing coding knowledge. In addition to building prototypes and websites, its GPT Engineer can ship fully functional web apps. It now claims to have 500,000 users who are building over 25,000 new products daily.."
Results are in for the L&D Global Sentiment Survey 2025, the 12th annual check of the pulse of L&D worldwide: “The L&D Global Sentiment Survey, now in its 12th year, once again asked two key questions of L&D professionals worldwide:
What will be hot in workplace learning in 2025?
What are your L&D challenges in 2025?
For the obligatory question on what they considered 'hot' topics, respondents voted for one to three of 15 suggested options, plus a free text ‘Other’ option. Over 3,000 voters participated from nearly 100 countries. 85% shared their challenges for 2025.”
Here’s my question - Don Taylor et al put a ton of work into this survey and clearly a lot of you responded, so now what will you do with the information? How will you act on it?
MEETING DELEGATE: Benchmarking LLMs on Attending Meetings on Our Behalf: So is the new joke going to be “this meeting could’ve been an AI meeting? > > “In contemporary workplaces, meetings are essential for exchanging ideas and ensuring team alignment but often face challenges such as time consumption, scheduling conflicts, and inefficient participation. Recent advancements in Large Language Models (LLMs) have demonstrated their strong capabilities in natural language generation and reasoning, prompting the question: can LLMs effectively delegate participants in meetings? To explore this, we develop a prototype LLM-powered meeting delegate system and create a comprehensive benchmark using real meeting transcripts.”
Ron Wyden asks for rules about whether you own your digital purchases: Can’t AMEN this loudly enough > > ‘To put it simply, prior to agreeing to any transaction, consumers should understand what they are paying for and what is guaranteed after the sale.’
Workhelix taps years of research to help enterprises figure out where to apply AI: This sounds both needed and slightly terrifying > > “Workhelix breaks down a company’s employee positions into specific job functions and tasks and scores each task for its suitability for AI adoption. This helps companies build roadmaps for how and where to adopt AI and gives enterprises a way to monitor if the AI they adopted is working.”
You.com launches a 'professional-grade' research agent for enterprises: This sounds really valuable but this is some next level corporate speak > > “ARI's breakthrough is its ability to maintain contextual understanding while processing hundreds of sources simultaneously," said Bryan McCann, co-founder and CTO of You.com. "When combined with chain-of-thought reasoning and extended test-time compute, ARI is able to discover and incorporate adjacent research areas dynamically as analysis progresses.” Another one: Tech mogul Mel Morris announces public launch of AI research engine Corpora.ai - “We’re a research engine — we’re not a search engine,” Morris said. “The breadth and depth of what we produce and look at is really important.” > > here’s the thing, these engines will always be limited by what data they can actually get at, e.g. the tons of data behind paywalls and internal corporate data that would lend context to a lot of this - the key piece here though are the models they’re building for being able to bring all this data together - so will you be trusting AI to go rummaging through every bit of corporate data so it can build the most nuanced context?
Backend game engine Pragma levels up player engagement tracking with FirstLook acquisition: This is one of those signals that I want you to not ignore because you don’t build games…this could be widely applicable > > “The best games and most successful aren’t just well-built; they’re supported by passionate communities, grown through word of mouth, virality, quality influencer engagements and more,” said Eden Chen, CEO of Pragma. “Bringing FirstLook into the Pragma ecosystem means studios no longer need to piece together fragmented solutions to engage with their players. Together, we’re bridging game infrastructure with community engagement, giving developers the tools to not only scale their games but also build thriving player communities from day one.”
ElevenLabs’ new speech-to-text model Scribe is here with highest accuracy rate so far (96.7% for English): Go ahead - try it here.
Fiber computer allows apparel to run apps and 'understand' the wearer: This is cool but what if your clothes were watching you? > > “What if the clothes you wear could care for your health? MIT researchers have developed an autonomous programmable computer in the form of an elastic fiber, which could monitor health conditions and physical activity, alerting the wearer to potential health risks in real-time.”
How Factory is turning AI into ‘a junior developer in a box’: Again, I know you’re reading this and going “I’m not a coder, I build courses” and I hear you but do you really think that if an AI can write code that it can’t create courses, maybe not as good as yours, but good enough? > > “The next step beyond that is AI coding assistance that’s more agentic—capable of handling at least certain tasks from start to finish without constant supervision. That’s what a San Francisco startup called Factory offers with its platform, which officially debuts today.”
Legal tech startup Luminance, backed by the late Mike Lynch, raises $75M: We’re getting there aren’t we? The more rule-bound your operating environment is, the more deeply it can disrupted by AI (once the AI is trained on those rules) > > “Luminance uses what it calls a “Panel of Judges” AI system to automate and augment a business’ approach to contracts — including generation, negotiation, and post-execution analysis. The startup uses a proprietary large language model (LLM) to power its main product, Lumi Go, which lets customers send draft agreements to a counterparty and have the AI auto-negotiate on their behalf.” > > This is a key read for another reason - the “panel of judges” model - what do you know about it? What does it compare to? In what environments does it perform best? How much should pay for it? These are questions that an org will have to develop the expertise to answer and not just the CIO - wouldn’t be cool if there were some team that focused on learning, that could help with that?
The Labor Market Effects of Generative Artificial Intelligence: An important one to revisit, especially as we get further down this road > > “We find, consistent with other surveys that Generative AI tools like large language models (LLMs) are most commonly used in the labor force by younger individuals, more highly educated individuals, higher income individuals, and those in particular industries such as customer service, marketing and information technology. Overall, we find that as of December 2024, 30.1% of survey respondents above 18 have used Generative AI at work since Generative AI tools became public.” Another labor market study: Is AI already shaking up the labor market? Four trends point to major change > > “The study measures more than 100 years of "occupational churn"—or each profession's share in the U.S. labor market—for a historical look at technological disruption. It revealed a stretch of stability between 1990 and 2017 that runs counter to popular narratives about robots stealing American jobs.”
Edinburgh history: 3 million records of Scotland's past now available with just a click: “It combines designation records of Scotland’s most important historic sites and buildings,; the catalogue of Scotland’s archaeology and built heritage,; more than 2000 culturally significant objects from HES’s properties; and educational and archival images and media from SCRAN.”
Hate #vibecoding? Try vibecoding by voice!
https://addyo.substack.com/p/speech-to-code-vibe-coding-with-voice?utm_source=post-banner&utm_medium=web&utm_campaign=posts-open-in-app&triedRedirect=true