0:00 T3 Code just shipped in-app visualization capabilities.0:03 That means an AI agent can now build a dynamic, interactive experience for you right inside your conversation thread.0:12 Last week, we talked about AI moving from demos to industry.0:16 This is what that looks like in practice.0:19 The ground is shifting from what a model can say to what an agent can do.0:24 The theme this week is simple.0:26 The conversation among builders is no longer about the raw intelligence of the model.0:32 That's table stakes.0:33 The new frontier is usability.0:35 The new obsession is the workflow.0:38 How do you make these powerful agents not just smart, but genuinely useful, day-to-day?0:44 We're seeing the scaffolding for that future go up right now.0:48 Here are the headlines that matter.0:51 First, that T3 Code update.0:53 Theo, the CEO, announced that agents can now build visualizations directly in a thread.0:59 He gives a shoutout to the engineer who pushed for it, saying he didn't expect to "dig it so much." That's the signal.1:07 Even the people building it are being surprised by how powerful it is to move beyond text.1:13 This isn't just about making things look pretty.1:17 It’s about changing the nature of the interaction.1:20 Second, a contributor at Nous Research dropped a new resource for the Hermes Agent.1:26 People have been asking which "skills" to use with it.1:30 So they've compiled the answer into a YouTube playlist.1:34 It covers which skills to install and—this is key—how to write your own.1:39 This is about extensibility.1:41 It's the move from a closed box to an open ecosystem where users can add the specific capabilities they need.1:49 It’s the app store moment for agents.1:51 And third, a post that went viral points to a twenty-eight-minute video from the engineer who built Claude Code.1:59 It's a deep dive on how to write prompts that actually work.2:03 Not hype, not magic phrases.2:05 It’s about the mechanics.2:07 CLAUDE dot md files, memory shortcuts, parallel sessions, specific prompting techniques.2:13 The post notes that three-hundred-dollar courses don't even touch what this engineer shows for free in the first ten minutes.2:22 This is about user mastery.2:24 It's about closing the gap between what the tool can do and what the user knows how to do.2:30 So what does it all add up to?2:32 You have a new kind of interface.2:35 A way to add new capabilities.2:37 And a guide to becoming a power user.2:39 Separately, they're just updates.2:42 Together, they're a pattern.2:44 They are the three legs of the stool for making AI agents a real, productive platform.2:50 And that's where we're going to dive deep.2:53 Because this isn't about one company or one feature.2:56 This is about a fundamental change in how we interact with computation.3:01 Okay.3:02 Let's unpack this.3:03 The real story of the week isn't in any single announcement.3:07 It's in the convergence.3:09 For months, the public conversation about AI has been dominated by the models themselves.3:15 Is GPT-5 coming?3:16 How many parameters does this new open-source model have?3:20 Which model is better at writing poetry?3:23 That was the game.3:24 A race for raw intelligence, measured by leaderboards and benchmarks.3:29 That race isn't over.3:31 But it's no longer the ONLY race.3:33 A new one has started.3:35 And it's not about building a bigger brain.3:38 It's about building a better car around the engine.3:41 It's about the dashboard, the steering wheel, the pedals.3:45 It's about the entire user experience.3:48 This week, we saw three critical pieces of that experience get a major upgrade, out in the open.3:55 Let's start with the interface.3:57 T3 Code.3:57 When Theo says they shipped "in-app visualization capabilities," it sounds dry.4:03 It sounds like a feature for a changelog.4:06 It is NOT.4:07 This is a paradigm shift hiding in plain sight.4:10 Think about how you interact with an agent right now.4:14 You type a question.4:15 It types back an answer.4:17 Maybe it gives you a block of code.4:19 Maybe it writes a long document.4:21 But it's all text.4:23 It's a transcript.4:24 You, the human, have to take that text, parse it, understand it, and then go do something with it in another application.4:32 If you ask for sales data, you get a table made of text characters.4:37 You then have to copy that, paste it into a spreadsheet, and build your own chart.4:43 What T3 Code just did is collapse that entire workflow.4:47 The agent doesn't just give you the data.4:50 It can now build the chart for you, right there in the conversation.4:54 It's not a static image of a chart.4:57 It's a "dynamic experience." That means an interactive component.5:02 You could presumably mouse over it, filter it, change the date range.5:06 The conversation thread is no longer just a log of text.5:10 It's becoming a canvas.5:12 An application host.5:13 Think about the implications.5:15 You're a developer debugging a complex piece of software.5:19 Instead of the agent describing the data flow, it just renders a live, interactive diagram of it.5:26 You click on a node, and the conversation continues from there.5:31 You're a financial analyst.5:33 You ask the agent to model a few scenarios.5:36 Instead of getting three massive walls of text, you get an interactive slider.5:41 You drag the slider for the interest rate, and the revenue projection updates in real time.5:48 This is HUGE.5:48 It dissolves the boundary between the AI and the tools you use.5:53 The AI is not just a consultant you talk to anymore.5:57 It's a builder that works alongside you, in the same space.6:01 It turns the chat window from a command line into a graphical user interface that's being created on the fly, just for you, based on your needs at that exact moment.6:12 Theo's comment is telling.6:14 "Didn't expect to dig it so much." That's the feeling of stumbling into the future.6:20 It sounds like a small feature, but it feels like a new medium.6:24 It makes every other text-only interface feel instantly archaic.6:29 It’s the difference between reading a book about flying and actually being in the cockpit.6:35 This is the first pillar of agent usability: a rich, dynamic, and integrated interface.6:41 So you have a better interface.6:44 The agent can show you things, not just tell you.6:47 But what can the agent actually DO?6:49 What can it connect to?6:51 That brings us to the second pillar: extensibility.6:55 And that’s the story of the Hermes Agent from Nous Research.6:59 The post from the contributor 'witcheer' is subtle but important.7:03 They say, "we see a lot of questions about which skills people use with Hermes Agent." Stop right there.7:11 That question itself is the signal.7:13 People aren't asking "what is Hermes Agent?" They aren't asking "is it smart?" They are asking "what skills should I install?" This implies a user base that has already moved to the next level of thinking.7:28 They understand the base model is just a starting point.7:32 The real power comes from connecting it to other tools and data sources.7:37 A "skill" is just that—a piece of code that lets the agent interact with an API, a database, or another service.7:45 It's the agent's equivalent of hands and feet.7:48 By releasing a YouTube playlist on "which skills to install, and how to write your own," Nous Research is doing something profound.7:57 They're not just shipping a product.8:00 They are cultivating an ecosystem.8:02 They are teaching people how to fish.8:05 This is the classic platform playbook.8:07 Think of the iPhone.8:09 When it first launched, it was a beautiful, closed box.8:13 It had a handful of Apple-made apps.8:15 The REVOLUTION was the App Store.8:17 When Apple allowed third-party developers to build their own apps, the iPhone's capabilities exploded.8:25 Suddenly, it wasn't just what Apple thought you should do with your phone.8:30 It was what a million developers around the world could imagine.8:34 That's what's happening here with agent skills.8:38 An agent with no skills is just a brain in a jar.8:41 It can think profound thoughts, but it can't order a pizza.8:45 It can't check your calendar.8:47 It can't post to your team's Slack channel.8:50 A skill for each of those things is what makes the agent practical.8:55 And the fact that the guide covers how to write your OWN skills is the most critical part.9:01 It means you're not limited to some pre-approved list in a marketplace.9:06 If your company uses a bespoke internal tool for customer support tickets, you can write a skill that lets your agent read and file those tickets.9:17 The agent becomes a true member of your team, integrated into your specific, unique workflows.9:23 This is the answer to the scaling problem.9:26 You don't need one monolithic AI model that knows everything about every possible API in the world.9:33 That's brittle and impossible to maintain.9:36 Instead, you have a lean, smart core model, and a library of skills that you can plug in as needed.9:43 It's modular.9:44 It's decentralized.9:45 And it puts the power in the hands of the user.9:49 So now we have two pillars.9:51 A dynamic interface where the agent can build experiences for you.9:55 And an extensible skill system so the agent can take action in the world.10:00 That's a powerful combination.10:02 But there's a missing piece.10:04 You.10:05 The user.10:05 This brings us to the third and final pillar: user mastery.10:09 And this is where that twenty-eight-minute video from the Claude Code engineer comes in.10:16 The post sharing it frames it perfectly.10:18 The engineer who BUILT the tool is showing you how to use it.10:23 For free.10:23 And the techniques are so potent that they make three-hundred-dollar paid courses look like a joke.10:30 What does the video cover?10:32 "CLAUDE dot md files, memory shortcuts, parallel sessions, techniques of prompting." Let's break that down.10:40 This isn't fluffy, abstract advice like "be clear and specific." This is tactical, operational knowledge.10:47 A CLAUDE dot md file is likely a way to give the agent persistent context and instructions for a project, like a constitution but more integrated.10:57 Memory shortcuts suggest ways to efficiently manage the agent's limited attention window.11:04 Parallel sessions imply running multiple independent conversations at once to tackle a complex problem from different angles.11:12 These are power-user features.11:15 This is the stuff that separates an amateur from a professional.11:19 This is the most overlooked part of any technological revolution.11:24 It's not enough to have a powerful tool.11:26 The user has to know how to wield it.11:29 A Formula One car in the hands of a brand-new driver is slow and dangerous.11:34 The same car, with a professional at the wheel, is a miracle of engineering.11:40 For the past year, "prompt engineering" has been the term of art.11:44 But a lot of it has been a kind of digital alchemy, with people sharing "magic" prompts that supposedly unlock the AI's secret powers.11:54 Most of it was noise.11:55 What this video represents is the shift from alchemy to chemistry.12:00 From superstition to science.12:02 The engineer isn't giving you magic words.12:05 They're explaining the underlying mechanics of the system.12:09 They're giving you a mental model of how the agent thinks, how it remembers, and how it processes instructions.12:17 When you understand the mechanics, you don't need magic words anymore.12:22 You can construct the right prompt, the right workflow, for any situation, from first principles.12:29 This is how you close the capability-to-utility gap.12:32 This is how you make sure the incredible power of these new tools doesn't go to waste.12:38 You educate the user.12:40 Not with marketing, but with real, substantive, operational training, straight from the source.12:46 So, let's put it all together.12:49 This week wasn't about a new, bigger, smarter model.12:52 It was about making the models we already have actually work for us.12:57 It was about building the rest of the stack.13:00 T3 Code showed us the future of the interface: dynamic, visual, and integrated into the conversation.13:07 Nous Research and the Hermes Agent showed us the future of capability: modular, extensible, and community-driven through skills.13:16 And the Claude Code engineer showed us the future of the user: educated, empowered, and moving from a prompt guesser to a systems operator.13:26 The interface.13:27 The skills.13:28 The user.13:28 That’s the trifecta.13:30 That is the stack that turns AI from a fascinating curiosity into a productive, world-changing platform.13:37 The big story is that all three of these layers are being built right now, at the same time.13:44 The conversation has moved.13:45 It's no longer just about the engine.13:48 It's about building the whole car.13:50 And this week, we got a clear look at the blueprints.13:54 What this sets up is a new kind of competition.13:58 The platforms that win won't just be the ones with the highest benchmark scores.14:03 They will be the ones with the most fluid interfaces, the most vibrant skill ecosystems, and the most sophisticated users.14:12 They will win on usability.14:14 They will win on workflow.14:15 They will win by making their users smarter, faster, and more capable.14:20 The race for raw intelligence is heating up.14:23 But the race to build a truly usable agent platform has just begun.14:28 And that's where all the action is going to be.
