0:00 On March thirty-first, Salesforce turned Slackbot into an agentic desktop assistant that can monitor your screen and transcribe your meetings.0:10 This isn't just an upgrade to a chatbot.0:13 It’s a transformation of Slack into a proactive operating system for your work, powered by Anthropic's Claude.0:21 With over thirty new capabilities, it can now watch what you do, understand the context of your tasks, and execute workflows across third-party tools.0:32 This move signals a fundamental shift.0:35 AI is no longer just a feature within an application.0:39 It's becoming the application itself.0:41 That was the biggest move, but not the only one.0:45 Google just released Gemma 4, a suite of open models under an Apache 2.0 license, designed to run on everything from a mobile phone to a high-end GPU.0:56 The goal is accessibility.0:58 The largest model, a 31-billion parameter version, now ranks third on the Arena AI text leaderboard, placing it among the most capable open models available.1:09 But the real story is the smaller models, designed to run efficiently on consumer hardware like Android phones and even a Raspberry Pi.1:19 Google is trying to put a capable model in every developer's hands, regardless of their hardware budget.1:27 Meanwhile, a Chinese startup called MiniMax just launched MiniMax-M1.1:32 It's the world’s first open-source, large-scale hybrid-attention reasoning model.1:38 It supports a one million token context window and can output eighty thousand tokens of pure reasoning.1:45 What matters here is the efficiency.1:48 Using a new mechanism called Lightning Attention, it performs deep reasoning tasks with only thirty percent of the computing power of a comparable model like DeepSeek R1.2:01 The entire reinforcement learning training phase cost just over half a million dollars.2:07 That’s an order of magnitude less than expected, and it shows the barrier to creating state-of-the-art models is falling faster than anyone predicted.2:18 The money is also flowing toward specialized applications.2:22 Sona, a platform for frontline workforce management, just raised forty-five million dollars in a Series B round.2:30 Their goal is to replace the legacy software that manages scheduling, payroll, and operations for industries like retail and hospitality.2:41 As their CEO put it, every other software category has been transformed by AI, but the tools for the world's largest workforce are two decades old.2:51 Sona is betting that AI-native infrastructure is the only way to fix it.2:57 In that same vein, a company called Titan launched banking-native AI models.3:02 These are not general-purpose LLMs.3:05 They are tailored specifically for compliance and regulatory environments.3:10 The models have banking logic and regulatory frameworks embedded directly into their architecture.3:17 This allows for traceable reasoning, which is critical for audits.3:22 On their own benchmarks for banking tasks, they achieve seventy-six percent answer accuracy, outperforming general models that don't understand the constraints of a regulated industry.3:36 And finally, as all these new agents and models get deployed, the question of safety becomes urgent.3:43 An AI infrastructure startup named Vijil just raised seventeen million dollars to address this.3:50 Their platform is designed to enhance the trust, security, and resilience of enterprise AI agents once they are in production.3:59 It continuously evaluates them for reliability and safety, enforcing governance controls and runtime protections.4:08 It’s the unglamorous but essential plumbing required to make any of this enterprise-ready.4:14 Let’s go back to Slackbot.4:16 The change from a reactive bot to a proactive agent is the real story here.4:22 "Agentic" means it has autonomy.4:24 It can see you’re trying to schedule a meeting across three different channels, recognize the pattern, and offer to complete the task for you.4:34 It can join your Zoom or Google Meet call, transcribe the entire conversation, summarize it, and assign action items to the right people.4:44 It becomes a central hub that coordinates work across the six thousand applications in the Salesforce ecosystem.4:53 This is the promise of automating repetitive cognitive labor.4:57 Here’s the turn.4:58 To do this, Slackbot needs to see everything.5:02 The new capabilities include monitoring your desktop activity.5:06 It needs to know what applications you’re using and what data you’re looking at to understand your workflow.5:14 It needs access to your audio to transcribe meetings.5:18 This introduces a profound privacy trade-off.5:21 The convenience of an AI assistant that anticipates your needs is paid for with persistent, low-level surveillance of your work.5:31 We’ve seen this debate with consumer products, but this is one of the first major deployments of an always-on agent in a corporate environment.5:41 User acceptance of this feature will be a major test case for the entire industry.5:47 It’s no longer just about whether the AI is useful.5:51 It’s about whether we're comfortable with what it has to see to be useful.5:56 Now, let's look at Google's Gemma 4.5:59 The release of a highly capable open model is significant, but the strategy behind it is what matters.6:07 Google isn't just releasing one model; it's releasing a family of them, optimized for a spectrum of hardware.6:15 This is a deliberate ecosystem play.6:17 The smaller E2B and E4B models are built to run on-device, on mobile phones and IoT hardware.6:24 This is about distribution at massive scale.6:27 By making them efficient on consumer devices, Google is trying to make Gemma the default choice for any developer building an app that needs local, on-device intelligence.6:40 Then you have the high end.6:42 The thirty-one billion parameter model is a direct competitor to other top-tier open models, but it performs best on high-end NVIDIA H100 GPUs.6:52 So Google is bracketing the market.6:55 They are providing a solution for the independent developer tinkering on a Raspberry Pi, and a solution for the well-funded startup with a cluster of GPUs.7:06 The goal isn't just to have the best model.7:09 The goal is to make Gemma the most available model.7:13 This is Google attempting to build the foundational layer for the next generation of AI applications, much like Android became the foundational layer for mobile.7:25 They are betting that ubiquity will ultimately be more powerful than having the single highest benchmark score.7:33 We’ve stopped building applications with AI.7:36 We’re now building them on AI.
