AI costs are collapsing, models are specializing, and data centers are rewriting the rules of demand.
In this week’s Bricks, Bucks & Bytes episode:
🚨 AI token costs have collapsed, from $32 to $0.09 per million (race to zero begins)
🏗️ Data center demand is exploding, but electricians are the new bottleneck
🧠 “Specialized AI” is rising, but is it real value or just marketing spin?
But that’s not all:
Patric Hellermann questions whether AI “specialization” is just dressed-up commoditization.
Dustin DeVan breaks down why deterministic systems, not probabilistic AI, will win in enterprise.
Martin drops the uncomfortable truth: 90% accuracy might be useless if you're selling to professionals.
And then things get philosophical… Would you board a plane flown entirely by AI?
Founder Spotlight: Anna Berger (Trayd)
Raised a $10M Series A… in just 3 weeks. Moving fast in one of construction’s most overlooked bottlenecks: workforce + operations. We dig into:
What it actually takes to raise at speed in this market
Why construction labor + ops is still massively underserved
How Trayd is positioning itself in a crowded, noisy “future of work” narrative
What most founders get wrong about building in construction
🎧 Listen now for:
Why falling AI costs might destroy moats, not create them
The real constraint behind the data center boom (hint: it’s not capital)
Why enterprise AI needs governance, not just intelligence
The brutal truth about selling “90% accurate” tools in construction
How Anna Berger is scaling fast and where she’s placing her bets