0:00 Forty-six percent of all code written by GitHub Copilot users is now generated by AI.0:06 That number is the quiet beginning of a revolution that just went loud.0:12 This week, we saw the conversation shift from what AI can say, to what AI can do—autonomously, on your machine, without you watching.0:22 We're here to cut through the noise.0:25 Aaron Levie, the CEO of Box, tweeted this week about the "insane" amount of hype and BS around enterprise AI.0:33 He’s not wrong.0:35 But underneath that noise, a few signals are getting clearer.0:39 They point to a fundamental change in the architecture of work itself.0:45 We’re moving from the era of the chatbot to the era of the agent.0:50 Let's run through the headlines that are shaping this new reality.0:55 First, the platform ground is shifting.0:58 In a move that sent ripples through the developer community, Claude AI ended all third-party subscriptions this month.1:07 This wasn't a minor policy change.1:10 This broke hundreds of AI agents and automation tools that were built on top of its API, forcing a scramble to rebuild or pivot.1:20 It's a stark reminder that the ground you build on can be pulled out from under you.1:26 When a foundational layer becomes unstable, the entire ecosystem has to re-evaluate its dependencies.1:34 Meanwhile, Elon Musk’s Grok AI is leaning into a very different philosophy.1:40 They just rolled out version 4.20 with new speed and quality modes, giving users more granular control.1:48 But the real story was a viral exchange Musk amplified.1:52 A user tried to force Grok into a series of yes-or-no answers.1:57 The AI refused.1:58 It replied, quote: "Your list demands yes/no binaries that misframe my criteria...2:05 I reject the forced choice: substance over format." Musk's endorsement of this—of an AI prioritizing nuanced, evidence-based responses over simplistic binaries—signals a deliberate design choice.2:20 It’s being built for truth-seeking, not for appeasement.2:24 This is a direct challenge to models that prioritize safety filters to the point of losing factual integrity.2:33 Microsoft is also making moves to diversify its own foundation.2:38 After years of its AI strategy being almost synonymous with its partnership with OpenAI, Microsoft just launched three of its own new AI models in early April.2:50 This isn't just another product release.2:53 This is a strategic declaration of independence, a signal that Microsoft is building its own proprietary capabilities from the ground up, moving beyond just being the biggest customer of another company's tech.3:10 And while the foundational models compete, the application layer is where the work gets done.3:17 Salesforce just upgraded Slackbot into what it calls an autonomous work assistant.3:23 This is not the Slackbot you remember.3:26 With thirty new AI features, it now coordinates multi-step tasks and integrates directly with your CRM data.3:35 It’s a move to transform Slack from a chat app into the central nervous system for your entire business, powered by an AI that does more than just notify—it acts.3:47 Underneath all of this, the infrastructure costs are changing dramatically.3:53 Google announced new TurboQuant technology that cuts memory costs for running these models by a factor of six.4:02 This is not a small optimization.4:04 A six-X cost reduction in a core component of AI infrastructure is the kind of shift that enables new categories of applications to become economically viable.4:17 It makes deploying powerful AI at scale dramatically cheaper.4:22 Finally, the money is following the momentum.4:25 Elad Gil, a prominent investor and analyst, noted that both OpenAI and Anthropic are now rumored to be at a thirty-billion-dollar annual revenue run rate.4:37 Each.4:37 That’s zero-point-one percent of U.S.4:40 GDP...4:41 each.4:41 He projects that AI-related revenues could hit one percent of the entire U.S.4:47 economy by the end of this year, 2026.4:50 The numbers are starting to move from the abstract to the un-ignorable.4:56 So.4:56 We have platform instability, philosophical divergence in model design, and a massive re-platforming of work itself.5:05 But the single most important development connects all of these threads.5:11 It’s the story of how AI just got a job.5:14 Between January and April of this year, Anthropic didn't just update its model.5:20 It shipped a complete AI coworker stack.5:23 This is the change that matters most.5:26 The race, as one analyst put it, isn't about who has the smartest model anymore.5:32 It’s about who builds the best harness for AI to actually do work.5:37 Anthropic’s harness has two parts.5:40 The first is called Claude Cowork.5:43 This is not a chatbot in a browser window.5:46 This is a desktop agent.5:48 It has what the developers are calling "hands"—a sandboxed environment on your computer where it can read files, edit them, create spreadsheets, write formulas, and organize data.6:02 It can take a messy folder of receipts, read them, and generate an expense report in Excel, complete with SUM functions.6:11 It can take a transcript from a meeting and create a project plan in whatever app you use.6:18 This is the jump from conversational AI to autonomous execution.6:23 The previous generation of AI could suggest the Python code to perform a task.6:29 This generation just...6:31 does the task.6:32 It operates your software for you.6:35 The second part of the stack is an app called Dispatch.6:39 This is the remote control.6:41 It lets you send tasks to your desktop AI agent from your phone.6:46 So you can be on the train home, get a request, and just type into Dispatch: "Take the sales report from Q1, cross-reference it with the new client list from Salesforce, and create a presentation summarizing the top ten growth accounts.7:04 Put it on my desktop." And when you get home, the file is there.7:09 This three-part architecture—the Brain, which is the model; the Hands, which is the sandboxed execution environment; and the Session, which is the persistent memory of your work—is quickly becoming the standard for all AI agents.7:27 The MCP protocol, the framework that connects these agents to platforms, just hit ninety-seven million monthly SDK downloads.7:37 The adoption is happening at a blistering pace.7:41 And this isn't just Anthropic.7:43 The Salesforce upgrade to Slackbot we mentioned?7:47 It’s the same pattern.7:49 It’s about giving AI the permission and the tools to perform multi-step workflows across different applications.7:57 It’s about turning a notification tool into an autonomous assistant.8:03 We see the same pattern in the world of software development, which is often the canary in the coal mine for broader workplace trends.8:13 That opening statistic—forty-six percent of code on GitHub Copilot now being AI-generated—is astounding.8:21 But the more telling number is this: fully autonomous AI commits on public GitHub repositories grew six-fold in the last six months.8:31 They now account for about five percent of all commits.8:36 Let that sink in.8:37 This is not a developer accepting a suggestion from an AI.8:42 This is an AI agent, on its own, checking out code, writing a new feature or fixing a bug, testing it, and committing it back to the project.8:53 Without a human in the loop for that specific action.8:57 The AI is now a credited contributor on the team.9:01 This is the industrialization of cognitive work.9:04 For the past year, the debate was about AI's potential, its intelligence, its alignment.9:11 That was the laboratory phase.9:13 We are now leaving the lab and entering the factory.9:18 The tools are no longer just for thinking and talking.9:22 They are for doing.9:23 And they are being deployed at a scale that is already changing how the digital world is built.9:31 This is the trend that connects everything.9:34 The reason Claude AI shutting down third-party APIs was so disruptive is because those developers were already building the first generation of these agents.9:46 The reason Google’s cost-cutting matters is because deploying millions of these autonomous coworkers will require massive, efficient infrastructure.9:58 The reason the revenue numbers are exploding is because businesses are starting to pay for outcomes, not just conversations.10:07 They're paying for the presentation that gets made, the code that gets written, the report that gets generated.10:16 The hype from last year was about replacing search.10:20 The reality of this year is about augmenting—and in some cases, automating—work itself.10:27 This raises enormous, difficult questions about job displacement, the kind that led the CEO of Verizon to warn about twenty to thirty percent unemployment in the coming years.10:41 And it raises questions about how we even measure this.10:45 Economists are still debating the "missing productivity" from the last wave of IT.10:51 Now, productivity isn't missing.10:54 It's showing up as an autonomous commit on GitHub, or a completed task from a Slackbot.11:00 The conversation is no longer hypothetical.11:04 The agents are being hired.11:06 The work is being done.11:08 The debate was about whether AI could think.11:11 The reality is about what AI can now build.
