Can you really hack an airplane? And if so, how do you test for it without grounding the fleet for a year?

In Episode 28 of Built This Week, Sam Nadler and Jordan Metzner sit down with Eero Salih, CTO of Syberian, to explore how AI is transforming cybersecurity for commercial aviation.

We start with a live demo — a flight ops cyber radar Sam built to surface real-time security risks across airline operations. Then Eero breaks down what Syberian actually does: building digital twins of aircraft systems to run risk assessments without ever touching the physical plane.

This is critical because traditional penetration testing would ground an aircraft for up to a year for recertification. Syberian's AI-powered approach analyzes over 100 technical documents to map every computer system on board — from avionics to entertainment to crew scheduling — and identify vulnerabilities before they become incidents.

We also discuss:

Why cyber attacks on aviation are now classified as safety threats

New 2026 regulations forcing airlines to comply with stricter cybersecurity standards

How small teams are replacing developers with AI agent managers

The tools Syberian uses: Claude Code, Windsurf, Anthropic, and Gemini

Why Google and Anthropic are rejecting ads while OpenAI explores them

An ex-Amazon exec who vibe-coded a full CRM replacement in 72 hours

No hype.
No theory.
Just what happens when you put AI in charge of protecting critical infrastructure.

New episodes every Friday.

================================================================================

TIMESTAMPS
--------------------------------------------------------------------------------
(0:00) Why you can't hack-test an airplane
(0:45) Episode 28 kickoff and guest introduction
(1:30) Live demo: Flight ops cyber radar dashboard
(3:00) Analyzing real-time security threats across airline systems
(4:30) What Syberian actually does (in plain English)
(6:00) Why physical penetration testing grounds planes for a year
(7:30) Using AI to build digital twins of aircraft systems
(8:15) Hiring managers, not developers — AI agents do the coding
(9:30) Tools of the trade: Claude Code, Windsurf, Anthropic, Gemini
(10:00) New 2026 aviation cybersecurity regulations explained
(11:00) How cyber attacks became classified as safety threats
(12:30) The ripple effects: baggage weight, fuel calculations, pilot tablets
(13:30) Who are Syberian's customers? Airlines, private jets, and more
(14:55) AI News: Google and Anthropic reject ads in chatbots
(16:30) Why Anthropic's no-ads stance matters for enterprise customers
(17:30) Amazon exec vibe-codes full CRM replacement in 72 hours
(18:30) Why vibe coding works for internal tools but not production
(19:15) Final thoughts and closing

================================================================================

LINKS SECTION
--------------------------------------------------------------------------------
Built This Week
New episodes every Friday

Jordan Metzner
https://x.com/mrjmetz

Sam Nadler
https://x.com/Gravino05

Can AI actually reduce cloud costs — or does it just create better dashboards?

In Episode 27 of Built This Week, Sam Nadler and Jordan Metzner are joined by Ben, CEO of Espresso AI, to break down a real production system that uses machine learning to actively optimize data warehouse compute in real time.

We walk through a live demo built specifically to expose hidden inefficiencies inside Snowflake and Databricks environments — from over-refreshing dashboards to duplicated queries and underutilized clusters. Then we go deep on how Espresso AI works under the hood: proxy-based routing, workload-aware ML models, and fine-grained compute orchestration that runs without changing application code.

This is not FinOps theater. This is AI actively rewriting how compute is allocated.

We also discuss:

• Why most teams overpay for convenience in the cloud
• How real-time query routing beats manual cost controls
• Where AI helps engineers — and where it absolutely does not
• The limits of vibe coding for serious infrastructure
• Gemini powering Siri and what it means for voice assistants
• Meta’s massive GPU buildout and the future of hyperscalers

No hype.
No theory.
Just what happens when you put AI in control of real infrastructure.

New episodes every Friday.

Timestamps

(0:00) Why modern AI understands code differently
(0:45) Episode 27 kickoff and guest introduction
(1:30) Live demo: diagnosing hidden warehouse inefficiencies
(3:00) Why dashboards refresh far more than they are viewed
(4:30) The real cost of duplicated queries across teams
(6:00) What Espresso AI actually does (in plain English)
(7:45) Kubernetes for data warehouses, powered by ML
(9:30) How real-time query routing works
(11:30) Why most companies are not “doing it wrong”
(13:00) Transformers and deep code understanding
(15:00) Where AI helps engineers today
(16:30) Why AI cannot yet run core infrastructure autonomously
(18:00) Productivity gains without replacing engineers
(19:30) Gemini, Siri, and the next generation of voice assistants
(21:00) Meta’s massive GPU investments explained
(23:00) Will Meta become a hyperscaler
(24:30) Final thoughts and closing

Links Section

Built This Week
New episodes every Friday

Jordan Metzner
https://x.com/mrjmetz

Sam Nadler
https://x.com/Gravino05

Espresso AI
https://espresso.ai

Can AI actually beat prediction markets — or does the house always win?

In Episode 26 of Built This Week, Sam Nadler and Jordan Metzner kick off 2026 by breaking down a real AI trading bot Jordan built for prediction markets like Kalshi, using live market data, whale detection, coordination signals, and confidence scoring.

Jordan walks through the full system — backend, frontend, live alerts, and execution logic — and shares the honest results: a 66 percent win rate that still lost money once fees and market dynamics were factored in.

The takeaway is not hype. It is reality.

The episode also dives into:

• Why prediction markets feel like gambling but are regulated differently
• How insider-like signals emerge from coordination and volume behavior
• Why bots end up trading against bots
• Where real alpha might exist (and where it does not)

We also cover:

• Google NotebookLM as a serious education and onboarding tool
• Turning documents into infographics, slide decks, and audio learning
• Nvidia entering autonomous driving and competing with Tesla
• Nvidia’s new Rubin architecture and why it matters
• Tesla vs Waymo economics and the future of Full Self Driving
• Why Anthropic and Claude Code are becoming developer defaults

This is not theory.
This is what happens when you actually deploy AI systems into real markets.

Timestamps

(0:00) Why prediction markets are exploding
(1:07) Episode 26 kickoff
(2:00) Why build a trading bot at all
(4:30) Kalshi vs Polymarket APIs
(6:00) Live market signals and whale detection
(9:30) Win rate vs profitability
(12:00) Why fees destroy returns
(14:30) Bots trading against bots
(17:00) Where real alpha might exist
(18:00) NotebookLM for learning and onboarding
(21:00) Nvidia enters autonomous driving
(24:00) Tesla vs Waymo economics
(27:00) Nvidia Rubin chips explained
(28:30) Anthropic and Claude Code momentum
(29:30) Final thoughts

Links

Built This Week
New episodes every Friday

Jordan Metzner
https://x.com/mrjmetz

Sam Nadler
https://x.com/Gravino05