Weekly Link Roundup #60
To go anywhere, you have to start somewhere...
Because we all need some poetry in our lives > >
Marginalia by Billy Collins (excerpt) “And if you have managed to graduate from college, without ever having written “Man vs. Nature” in a margin, perhaps now is the time to take one step forward.”
Vibe Coding Tutorial: How to Build Web Apps Without Coding Using DeepSite: OK. I don’t even want to write anything about this. I just want you to go there. Enter a prompt - make up the wildest prompt you can think of for a Web site - or one that maps to something you want to see happen - then watch what this site cranks out in seconds. People, PEOPLE! When I say that the potential of AI needs to be personally experienced at all levels of the org (Yes! That means you Senior Leadership) this is what I mean. Until people puts hands on this tech, they just will not get that visceral experience of what it can do. Forget Hackathons - start some Try-a-thons. Lead the way #LearningAndDevelopment
Stanford professor on hybrid working and why companies should let employees choose: If you start talking to me about WFH or hybrid work, one of my first questions will be “how does that line up with any of Prof. Nick Bloom’s work? He’s the preeminent researcher on the impact and future of hybrid and remote work. This little article is key for one point - "What you see in the data is if you let people choose, that's best for their mental health.” Choice. Anathema to so many - imagine! Letting employees choose their work mode. Imagine having to do the work to create workplace structures around IT, mentorship, and innovation that can be flexible, reactive to to changing conditions and resilience. The companies that do will attract and retain the best talent.
The AI Agent Era Requires a New Kind of Game Theory: I’m going to leave this here so you can go ask your CISO about it. Think of it this way - every new LLM that you connect to your systems and every new agent that you add to that collection, can multiple attack surfaces and multiply the scale and speed that bad things can move at. Not spreading fear - just the need for awareness.
Swapping LLMs isn’t plug-and-play: Inside the hidden cost of model migration: This has been true in every enterprise system I’ve worked on. The cool thing here is the scale and speed with which things can break and change if you don’t plan accordingly. This is another reason my L&D should be taking the lead and raising the AI literacy bar across the org - you’ve got to know enough to at least ask the right questions. > > “In reality, each model interprets and responds to prompts differently, making the transition anything but seamless. Enterprise teams who treat model switching as a “plug-and-play” operation often grapple with unexpected regressions: broken outputs, ballooning token costs or shifts in reasoning quality.”
Claude just gained superpowers: Anthropic’s AI can now search your entire Google Workspace without you: The reason I’m including this is because this is a signal of what’s coming. Not just in Claude but across the market. That’s important because it will go further to eliminating what has been on your To Do list - what will you do with the time? > > “The new Research capability enables Claude to independently conduct multiple searches that build upon each other while determining what to investigate next. Simultaneously, the Google Workspace integration connects Claude to users’ emails, calendars, and documents, eliminating the need for manual uploads.” See also: Notion releases an AI-powered email client for Gmail.
Small Models, Big Impact: ServiceNow AI Releases Apriel-5B to Outperform Larger LLMs with Fewer Resources: Couple of things - #1 Good on ServiceNow. Models like this will become more and more important to enterprise deployments but also #2 Read that paragraph again. Could you explain the key points to sr leadership? Do you think sr leadership has been taught how to interpret those aspects? I’d argue that #LearningandDevelopment could work with IT (or whoever is bringing this into the org) to build learning experiences to make sure that everyone in the org can read that paragraph and grok it’s meaning. > > “To address these concerns, ServiceNow AI has released Apriel-5B, a new family of small language models designed with a focus on inference throughput, training efficiency, and cross-domain versatility. With 4.8 billion parameters, Apriel-5B is small enough to be deployed on modest hardware but still performs competitively on a range of instruction-following and reasoning tasks.”
Canva is now in the coding and spreadsheet business: Here’s why I love this - we can talk about competition in features and business models but I’m looking forward to the UI/UX competition. I think that AI will allow companies to focus more effort on differentiating their products at that level maybe even getting down to dynamic UIs > > “Canva is launching its latest smorgasbord of generative AI-powered workspace tools in its attempt to cram everything that Microsoft, Google, and Adobe can do into a single platform.”
Google launches Agent2Agent protocol to connect AI agents across platforms: Add this to MCP and AutoGen > > “Supported by over 50 technology partners, including Salesforce, SAP, ServiceNow, and MongoDB, A2A provides a universal framework for AI agents to securely exchange information, coordinate actions, and integrate across enterprise platforms.”
Reddit’s conversational AI search tool leverages Google Gemini: Remember back when Alexa and Siri came on the scene and people were like “voice is the new UI”? And all I could think of was how little I’d ever want to work in an office or be in a coffee shop when voice was the main UI. Turns out they were on the right track but its conversation and not voice that is the new UI > > “Google and Reddit announced the update on Wednesday, explaining that by incorporating Gemini on Vertex AI, it will help improve search relevance and provide quick answers for users.”
Meet Kairos: An AI Agent that Automates Workflows by Recording Your Screen: Tell me again how AI isn’t going to just feast on your To Do list > > “Here are some core aspects of Kairos:
Learns tasks by observing a user's screen activity and explanation.
It aims to automate the task after a single demonstration without requiring programming or visual workflow design.
It is designed to feel like showing a human co-worker how to do something.
Kairos turns the recorded actions into a repeatable automated process.
MIT study finds that AI doesn’t, in fact, have values: Is this more of AI mirroring us? > > “A study went viral several months ago for implying that, as AI becomes increasingly sophisticated, it develops “value systems” — systems that lead it to, for example, prioritize its own well-being over humans. A more recent paper out of MIT pours cold water on that hyperbolic notion, drawing the conclusion that AI doesn’t, in fact, hold any coherent values to speak of.”
The Shortlist for the International Booker Prize 2025 has Been Announced: Want to be better at innovation? Want more and more diverse ideas? Expand your thinking? READ FICTION. And if people ever try to shame you for spending time reading fiction (including my fav, Science Fiction), educate them on Sturgeon’s Revelation:
The Revelation: Ninety percent of everything is crud.
Corollary 1: The existence of immense quantities of trash in science fiction is admitted and it is regrettable; but it is no more unnatural than the existence of trash anywhere.
Corollary 2: The best science fiction is as good as the best fiction in any field.
Listening habits and subjective effects of background music in young adults with and without ADHD: Does your learning have a soundtrack? I just love how we tell kids to SIT STILL and BE QUIET when that may actually be hurting their ability to learn > > “Overall, the results confirm that most young adults listen to music during their daily activities and perceive positive effects from this listening.”
Stanford’s AI Index: 5 critical insights reshaping enterprise tech strategy: Great read >
The democratization of AI power is accelerating
The gap between AI adoption and value realization remains substantial
Specific business functions show stronger financial returns from AI
AI shows strong potential to equalize workforce performance
Responsible AI implementation remains an aspiration, not a reality
See also (from same report): The AI Race Has Gotten Crowded—and China Is Closing In on the US.
AI Education Is the New Space Race. Here’s How America Must Respond: Here’s the lede > > “China announced recently that all students in Beijing will receive mandatory training in artificial intelligence starting next fall.”
Issue #60 is da bomb! Many useful insights 😎