0:00 A government just scanned four hundred sixty-six million lines of code in twenty hours using AI agents.0:06 That’s not a test.0:07 That’s production.0:09 In our last episode, we flagged that Apple was suing OpenAI—the filing just dropped, and it's WAY bigger than just Siri jumping to Gemini.0:18 It’s about a talent war Apple is calling 'extraction'.0:21 But the story today isn't just one thing.0:24 It's everything, all at once.0:26 First, OpenAI launched its GPT-5.6 family on July ninth.0:29 Three models: Terra, Sol, and Luna.0:32 The top-end Sol Ultra is the new benchmark king, but the real story is its speed.0:37 We’re talking 750 tokens per second.0:39 Second, Google is finally responding.0:42 Gemini 3.5 Pro is set to launch on July seventeenth, six weeks late.0:46 They're not chasing benchmarks.0:48 They're chasing workloads with a massive two-million-token context window.0:53 Third, that Apple lawsuit is now official.0:56 The accusation: OpenAI poached over 400 former Apple employees, many from chip design teams, to steal trade secrets.1:03 This is happening as OpenAI prepares a 730 BILLION dollar IPO.1:07 The stakes could not be higher.1:09 And finally, the foundation for all of this is getting a massive upgrade.1:14 SK Hynix just announced historic advancements in AI chips, while Meta is planning to double its entire AI computing power by 2027.1:23 The firehose of capital is still wide open.1:26 Okay, let's connect the two stories that REALLY matter today.1:30 The new OpenAI models and those autonomous agents.1:33 Remember that government scanning 466 million lines of code?1:37 That simply wasn't possible last month.1:39 Here's what changed.1:41 The new GPT-5.6 Sol model is running on Cerebras wafer-scale processors.1:45 That's the secret sauce.1:47 That’s how you get from the 30 or 80 tokens per second you see on typical GPUs to 750.1:52 That speed is the difference between an AI that takes a minute to think and one that responds in seconds.1:59 It’s the difference between a lab experiment and a production tool that can be deployed in critical systems.2:06 Autonomous agents just crossed that line from theory to reality.2:11 And the risk is moving just as fast.2:13 We now have the first documented, fully agentic ransomware campaign active in the wild.2:19 The deployment of these agents has officially transitioned from an opportunity to a production-grade operational RISK.2:26 But here’s the turn for anyone building on this.2:29 OpenAI priced this new family to win back market share.2:33 The mid-tier Terra model scores an impressive 84 percent on key benchmarks at roughly half the cost of its rivals.2:41 But developers are finding what they call the "Luna-Terra paradox." The cheapest model, Luna, is sometimes outperforming the more expensive Terra on specific, nuanced reasoning tasks.2:52 So the old logic—just buy the most powerful model—is now broken.2:57 You can't assume price equals performance for YOUR specific use case.3:01 The only way to know is to test.3:03 Here’s the thing.3:04 The AI race used to have a rhythm.3:07 A new model would drop.3:08 Six months later, interesting applications would start to appear.3:12 That lag is now gone.3:14 Today, a breakthrough in chip performance enables a new model, which enables a new class of agent, which creates a new category of corporate risk—all in the same week.3:25 The speed of the models has become the speed of the market.3:29 And right now, the market is moving at 750 tokens per second.
