The final Edge Bites of 2025 — and what a week to end on.

This episode: the rules of business are being rewritten across marketing, assets, and operations all at once. Amazon and Meta are reorganising their AI divisions (again). The Chief AI Officer is having its moment — but will it end like the Chief Digital Officer experiment? UK banks are trialling AI agents that move your money autonomously. And the gap between "companies citing AI layoffs" and "economists seeing no macro impact" is widening.

Case Study: Klarna fired 700 people, replaced them with AI, bragged about it — then quietly started hiring humans back. Gary Marcus calls it "The Klarna Effect." The lesson? They found the boundary between AI capability and customer expectation by crashing into it.

Cutting Edge: Google, DeepMind, and MIT tested 180 multi-agent configurations. The verdict? "More agents is all you need" is wrong. On sequential tasks, every multi-agent system made things worse — some by 70%.

Tool of the Week: Arthur Bench — an open-source framework to benchmark AI models on your actual data. Stop vibe-checking. Start measuring.

Thanks for listening in 2025. We're back mid-January 2026.

AI is splitting knowledge workers into two camps — and only one has a future. This week: three big ideas from the AI Advent calendar on LinkedIn, Accenture's 30,000-consultant bet on Anthropic, why only 6% of companies trust AI agents, Perplexity's multi-front distribution war (and the lawsuit that could sink it), plus research that flips the AI-jobs narrative. Your tip: stop torturing your EA with rambling voice notes.

Edge Bites - the sharp take on AI and business. Every Friday.

Right Here, Right Now.

Your board meets 10 times a year. Last month alone saw major AI releases from Google, OpenAI, Anthropic, Meta and X. That's not oversight—that's time-lapse.

This week: the widening gap between AI promise and reality. Sam Altman issues a "code red" over Google. Microsoft quietly lowers AI sales targets. And the $40 million startup that claimed 90% AI automation—when the real number was zero.

Plus: self-teaching AI that needs no human training data. What it means for your investments.

10 minutes. One edge. Let's go.

Companies are slashing managers and calling it transformation. We're calling it a mistake. This week: why the smartest competitors evolve management, not eliminate it. Plus: Mastercard's playbook, research showing AI can't out-think grad students, and a free leadership tool.

10 minutes. One edge. Every week.

Mid-market companies face an impossible challenge: fighting resource-rich giants above while fending off nimble startups below. This week's Edge Bites breaks down the gap exploitation framework that lets you win on both fronts. Plus: the model wars prove AI's commoditizing, surprising research on generational AI benefits, Microsoft vs Adobe's opposing bets, and how to automate competitive intelligence at scale.

10 minutes. All the bites that you need at the edge.

Everyone's selling you the same lie: move fast or get left behind. But with 80% of AI projects failing, speed isn't your problem—constraints are.

This week: Nvidia's $800B single-day wipeout and what it signals, why AI costs dropping 280x changes everything, how Graphic Packaging turned $5M into $26M by solving constraints first, Netflix's research proving 93% of recommendation value comes from targeting not exposure, plus two AI calendar tools worth testing.

The VORTEX framework for mapping what's actually holding you back. Because smart constraints beat raw speed every time.

10 minutes. One edge. Every week.

This week: the uncomfortable truth about 10x growth. When technology improves tenfold, someone loses control — and they push back hard.

We're watching it happen now. Tech markets questioning AI valuations. Getty losing copyright battles it thought were settled. Amazon cutting 14,000 jobs while claiming "AI lets us do more with less." And the ultimate irony: Amazon — the automation king — sending cease-and-desist letters to stop... automation.

Meanwhile, one company went from $2.8M to $150M in revenue in under a year. Not by optimizing what they did, but by hunting value where nobody else was looking. That's Replit's story, and it's a masterclass in asking better questions.

Plus: AI engineering teams that just made human engineers optional, and the prompt framework that 10x'd my research speed this week.

The lesson? Stop asking "how do we use AI cheaper?" Start asking "who else could we serve?"

10 minutes. One edge. Every week.

Your company just spent 18 months debating Claude versus ChatGPT. Multiple pilots running. Zero competitive advantage. Welcome to the $250 billion AI waste pile.

This week: why 95% of companies are selecting tools before defining strategy—and funding their competitors' success. Ray Dalio's bubble warning and why this is railway mania, not tulip mania. Nvidia's 100,000 robo-taxis signal AI is moving from screen to street.

The Société Générale disaster: 400 AI professionals, 300 use cases, tens of millions wasted. And the number that should terrify you: 45% of executives think AI is paying off. Only 27% of middle managers agree. That 18-point gap is your early warning system flashing red.

Ten minutes that might save you millions.

Why are companies lighting $250 billion on fire this year?

This week: the five myths destroying AI value (and how the 5% who succeed bust them), Coca-Cola's CFO-led playbook that actually works, why AI models lie systematically and safety tools can't detect it, small models beating giants, and "AI brain rot" making models dumber.

Plus: a brutally honest leadership feedback hack using AI.10 minutes. Zero BS. For executives who need results, not hype.

Subscribe to Executive Edge – the AI Edit at slateaxion.com

Why your AI cost-cutting strategy is about to backfire spectacularly.

This week: The problem with everyone deploying the same efficiency AI. A 12-person startup that made $500M breaking all the rules. The Bank of England pours cold water on the AI party. And, a hack that will get you posting to LinkedIn like a pro. Plus our new segment "Cutting Edge" — can AI agents develop gambling addiction? (Spoiler: the research says yes.)

10 minutes. One edge. No BS.

Edge Bites - the weekly podcast for executives who want results, not AI theatre.