About this episode Dive deep into the daily currents of AI with 'AI Daily: Real Shifts, Real Money—The Signals That Matter.' Each episode dissects the latest in new models, product launches, research breakthroughs, and funding rounds, sifting through the noise to highlight developments that genuinely reshape the landscape. Get expert insights that cut through the hype, equipping you with the knowledge to understand where AI is truly heading and what truly drives its evolution.
0:00 A three-person startup in Helsinki just raised four hundred million dollars.0:06 In episode 74, we talked about Nexus-Fusion and how new models are redefining diagnostics, but today the money is screaming a different story.0:17 The action is moving from the digital to the physical.0:22 That Helsinki startup is Koneisto AI.0:25 Their series A was led by Apex Ventures, and the valuation is a staggering two billion dollars.0:32 For a company with no product and no revenue.0:36 Here’s why.0:37 Koneisto isn’t building another large language model.0:41 They’re building brains for robots.0:44 Specifically, they claim to have cracked real-world robotic dexterity using a technique they call Causal Kinematics.0:54 Instead of learning from a trillion images of cats, their models learn cause-and-effect from physical interaction.1:03 A robot arm learns to pack a box not by mimicking a video, but by understanding the physics of gravity, friction, and object stability.1:14 The four hundred million dollars isn't a bet on software.1:18 It’s a bet that the most valuable AI in the next decade won't live in the cloud, but inside a machine that can pick, place, and assemble things in the messy, unpredictable real world.1:33 While Koneisto is building a fortress around its IP, Amazon just gave away the keys to its own kingdom.1:42 This morning, Amazon Web Services open-sourced Project Packer.1:47 This is the core logistics model that optimizes how every single Amazon fulfillment center packs its boxes.1:55 For years, this was one of their most guarded competitive advantages, saving them billions in shipping costs.2:04 So why give it away?2:06 It’s a classic move.2:08 They’re not being generous.2:10 They’re trying to make their internal tool the global standard.2:15 By open-sourcing Packer, they get thousands of outside developers to improve it, to build businesses on top of it, and to integrate it into every corner of the global supply chain.2:29 The more companies that use Packer, the more companies will need the specific AWS computing services that run it best.2:39 They are turning a competitive advantage into a utility they can tax.2:45 It’s a power play, shifting the battle from ‘our model is better’ to ‘everyone is using our model.’ It’s brilliant.2:54 And it’s ruthless.2:56 But the most important development today didn't come with a press release or a funding announcement.3:04 It came in a 12-page PDF uploaded to an academic server.3:08 Researchers at MIT just demonstrated a technique for what they call ‘targeted amnesia’ in neural networks.3:17 They can now surgically remove a specific piece of information from a massive, already-trained model.3:25 Without damaging the rest of its capabilities.3:29 Think about what that means.3:31 A model trained on copyrighted data?3:34 The lawyers call, you snip out the offending material.3:38 A facial recognition system with your photo?3:42 You file a request, and the model forgets it ever saw you.3:47 A state-sponsored AI trained on propaganda?3:50 A rival could, in theory, induce selective memory loss.3:55 This changes the entire conversation around data privacy and AI safety.4:00 Until today, the only way to remove data was to scrap the model and retrain from scratch, costing millions.4:09 Today, that changed.4:11 It’s now possible to edit an AI’s memory.4:14 So we have a four hundred million dollar bet on controlling robots in the physical world.4:21 A trillion-dollar company trying to control the logistics software of the entire planet.4:28 And a quiet lab in Cambridge that just figured out how to control the mind of the machine itself.4:36 The focus has shifted.4:38 The game is no longer just about building the most powerful intelligence.4:44 It’s about building the most obedient one.
