In this video
Talks about: The AI Trick That Cuts Your Costs by Ninety Percent
Most teams are lighting money on fire every time they hit an API.
The real insiders are swapping giant frontier models for tiny specialists that run faster, cheaper and closer to the user. The shift is so dramatic that entire workloads drop from cloud-scale bills to pocket change, and it unlocks a future where AI lives on your phone, your laptop and every blinking device around you.
In this episode we pull back the curtain on a simple idea with massive consequences. If you cluster every request an app makes, patterns appear. Those patterns reveal the truth. Most products do not need a trillion-parameter know-it-all.
The story goes further. Once a model becomes small and specialized, it can run on your iPhone, inside your browser or on any chip with a heartbeat. Suddenly AI moves from the cloud to the edge. Privacy increases. Latency disappears. Entire categories of products become possible.
Small models are not a downgrade. They are the business model everyone forgot to notice.
In this video
Talks about: AI Agents Can Pretend To Be You. Identity Is Now A Survival Problem
AI agents are getting good at one thing humans never expected: acting like us.
Not in a sci-fi way. In a “your calendar, your email, your tone, your habits” kind of way. And once an agent learns your patterns, it doesn’t just copy you. It multiplies. One becomes fifty. Fifty becomes an army. All moving faster than any security team can watch.
This is the part teams keep overlooking. Identity systems were built for passwords, roles, and predictable users. Agents don’t behave like that. They don’t stand still. They don’t log in once and stay in a neat box. They roam. They chain actions.
They make decisions inside your systems before anyone even notices.
Copilots sit inside codebases. LLMs read entire repositories. Developers lean on assistants that can refactor, rewrite, and annotate at a pace no human reviewer can track. And suddenly everyone is asking the same uneasy questions.
What did the model see?
What did it store?
Who is actually “acting” inside this system right now?
This is why identity is becoming the battleground of the AI era. Trust can’t be static anymore. It has to move with the agent. It has to understand context, timing, intent, and anomalies. It has to let humans stay in the loop without slowing everything down. And it has to be baked in from day one, not duct-taped in when things start catching fire.
Teams that treat identity like plumbing will get burned. Teams that treat it like strategy will survive the next wave.
In this video
Talks about: The Secret OS Running the Next Generation of Events
AI without context only makes noise. AI with context reshapes the room.
A new kind of operating system is emerging inside conferences and summits. Objectives get captured. Sponsors get mapped. Attendees get guided toward the one conversation that actually matters. The whole venue behaves like a living system that understands what people want and what the event is trying to achieve.
The shift begins with a simple idea. Old software needs every scenario prewritten. Real life does not. Agents can now make decisions in the moment as long as they understand who is in the room and what everyone is trying to do. The difference between a great outcome and pure chaos comes down to the quality of that context.
This episode opens the door to a world where identity, intent and goals become the building blocks for AI decisions. See how probabilistic thinking replaces rigid rules and how one contextual tweak can turn noise into real value for sponsors, speakers and attendees.
Context makes the difference between a useful nudge and a thousand meaningless pings.
