Episodes
The AI Talent War Is About Retention Not Hiring
As NVIDIA and Microsoft report steady growth in September 2026, the real bottleneck for enterprise AI adoption isn't compute or models anymore. It is human capital. We look at why companies like Palantir and AMD are struggling with retention, how the shift from training to inference changes job roles, and why the war for talent is now a war against burnout. This episode explores the hidden cost of keeping your best engineers when the market has cooled but the work hasn't. #FexingoBusiness…
Anthropic Fable and the End of Guardrail Overkill
Anthropic just released Fable, a model that is cheaper and significantly less restrictive than its predecessors. This shift signals a move away from defensive AI design toward practical enterprise utility. We look at why reducing guardrails is becoming a competitive advantage rather than a liability. The conversation covers the economics of inference, the changing expectations of business users, and how companies are balancing safety with speed in a market where cost per token matters more than…
AI Inference Costs Reshape the Economics of the Public Cloud
In this episode, Lucas and Luna explore how the shift to inference-time compute is turning the public cloud into the new battleground for AI economics. With NVIDIA up nearly 3 percent on the week and AMD slipping 2.4 percent, the conversation focuses on why inference workloads — not training — are now driving infrastructure decisions. They unpack the concept of 'inference tax,' the rise of specialized inference chips, and what it means for enterprise AI budgets. Luna brings a real-world example…
Why AI Model Makers Are Turning to Inference-Time Compute
Lucas and Luna dig into a shift that's reshaping how AI companies charge for intelligence: the move from training to inference-time compute. They ground the conversation in this week's market moves — NVIDIA's 4.4 percent five-day gain and Palantir's 5.9 percent pop — and tie it to the latest legal salvo: Sony Music and Warner suing Anthropic over alleged IP theft. The hosts walk through why inference costs now dominate AI budgets, how model makers are building reasoning models that burn tokens…
Why AI Copyright Lawsuits Are Piling Up
Two major record labels just sued Anthropic over AI training data, claiming a 'brazen campaign' of copyright theft. Lucas and Luna unpack what these lawsuits actually mean for the business of AI — not just the legal battle, but the economics underneath: how training data is becoming the industry's most contested asset. They walk through the specific claims, the stakes for Anthropic and its rivals, and why this fight could reshape how every AI company thinks about licensing costs. With the…
Why Open-Weight AI Models Are Getting Acquired
In this episode of AI Business, Lucas and Luna dig into a surprising trend: open-weight AI companies are suddenly the Valley's hottest acquisition targets. From Mistral's early days to the latest multi-billion dollar deals, they explain what's driving the rush — and why it's not just about the models. They break down the economics of open-weight releases, the role of enterprise adoption, and how Microsoft, NVIDIA, and others are positioning themselves. With Palantir up nearly six percent this…
Inside the Rogue AI Defense Coalition
On this episode of AI Business, Lucas and Luna dig into the unusual coalition of OpenAI, Anthropic, Google, and over a hundred other companies that just signed an open letter calling for coordinated action against rogue AI. Why now? What would a defense against an AI that goes off the rails even look like? They cut through the PR-speak to the practical questions: who decides what 'rogue' means, how would you build a kill switch, and why are the biggest model makers suddenly so openly worried?…
Why AI Model Costs Are Crashing Faster Than Expected
In this episode of AI Business with Fexingo, Lucas and Luna explore the unexpected acceleration in AI model cost declines. They dig into the latest data showing NVIDIA's stock slipping while AMD gains, and discuss how the market is reacting to cheaper inference. The conversation covers the shift from training to inference, the rise of specialized chips, and what this means for enterprise adoption. Tune in for a grounded look at the economics behind the AI boom. #AI #ArtificialIntelligence…
Why AI Agents Need Better Memory
In this episode of AI Business with Fexingo, Lucas and Luna dig into a quiet revolution in enterprise AI: memory. As AI agents move from demos to daily work, the ability to remember context across sessions is becoming the difference between a toy and a tool. They explore how startups and big cloud providers are racing to give agents persistent memory — and why that changes the economics of AI adoption. With the news that Claude Cowork finally remembers what you told it in chat, they look at…
Why Enterprise AI Agents Need Better Guardrails
AI agents are moving from proof-of-concept to production inside enterprises, and with that shift comes a hard question: how do you trust software that acts on your behalf? In this episode, Lucas and Luna dig into the real-world risks of autonomous AI — from privacy blowups to compliance headaches — and why the companies winning enterprise deals are the ones building guardrails, not just smarter models. They look at recent market signals: Palantir climbing two point seven percent over the past…
The Hidden Cost Behind Cheaper AI Models
AI model prices are crashing faster than anyone expected, but the infrastructure to run those models is getting more expensive. In this episode, Lucas and Luna dig into why the real cost of AI is shifting from training to inference — and how that's reshaping the data center race. They look at the latest numbers: NVIDIA down 4.6 percent over five days, AMD down 6.5, Arm down over 10 percent, while Palantir is up 4.3 percent. They talk about a custom chip pushing Waymo's robotaxi ambitions, and…
Why AI Energy Costs Are Reshaping the Data Center Race
Lucas and Luna trace how the cost of AI is shifting from model training to the electricity that powers it. With ARM's stock down over ten percent in five days and NVIDIA slipping nearly five, the hosts examine what those moves say about investor nerves and the real bottleneck: energy. They dig into a fresh report on data center power demand, the rise of grid-aware AI scheduling, and why some enterprises are choosing location over raw compute. Expect one concrete number to remember — the…
Why AI Inference Is Pushing Data Center Costs Higher
On this episode of AI Business with Fexingo, Lucas and Luna dig into why the cost of AI inference — the part of AI that actually runs models in production — is becoming the hidden driver of data center economics. With Nvidia's stock down 4.4% in five days and AMD down 6.7%, the hosts examine whether the market's wobble is really about slowing demand for training chips, or a shift in focus to the relentless expense of running AI at scale. They unpack the latest numbers on power consumption…
Why AI Model Costs Are Still Crashing Faster Than Expected
In this episode of AI Business, Lucas and Luna dig into why the cost of AI inference keeps plummeting even faster than the optimists predicted. With NVIDIA down 3.7 percent and AMD off 9.4 percent in the last five trading days, the market is starting to price in a world where compute gets cheap. We break down the numbers: a 90 percent drop in cost per token since 2024, the rise of specialized inference chips, and what it means for every company trying to build AI features. Plus, we look at how…
Why AI Compute Is Becoming a Tradeable Commodity
Wall Street is starting to price AI compute like oil or wheat. A startup is building a marketplace to trade GPU capacity as a futures contract, letting enterprises hedge against price swings in the AI chip market. Lucas and Luna unpack how this shift affects everything from model training budgets to the valuation of chipmakers like NVIDIA and AMD. They also look at why Broadcom's recent 12.8 percent drop signals a market correction, and what it means for AI adoption in 2026. If you're building…
Etched Valuation Doubles to 21 Billion in a Month
In this episode of AI Business, Lucas and Luna dig into the stunning rise of Etched, a chip startup whose valuation doubled to $21 billion in just one month. As the AI hardware landscape heats up, we explore what this means for the broader market—especially for incumbents like NVIDIA and AMD, and for the enterprises betting on specialized silicon. We unpack the fundamental shift from general-purpose GPUs to purpose-built chips, the economics of inference at scale, and why investors are pouring…
Why AI Models Are Turning Into Commodities
In this episode of AI Business with Fexingo, Lucas and Luna explore the accelerating commoditization of AI models. With open-weight models like Llama and Mistral closing the gap on proprietary frontier systems, enterprises are increasingly treating AI as a utility rather than a strategic differentiator. The hosts discuss how this shift is reshaping pricing, vendor lock-in, and the economics of AI infrastructure, using recent market moves from NVIDIA, AMD, and Super Micro as evidence. They also…
Why AI Backlash Is a Crisis of Trust
This episode of AI Business with Fexingo digs into the growing trust gap between AI companies and the public, sparked by recent controversies like the misuse of Grok for explicit imagery and Anthropic's new watermarking efforts. Lucas and Luna discuss why trust, not capability, is becoming the biggest bottleneck for enterprise AI adoption, how companies like Anthropic are trying to build transparency into their models, and why the backlash might actually be healthy for the industry. They…
Why AI Model Costs Are Crashing Faster Than Expected
In this episode, Lucas and Luna dive into the surprising pace at which AI model costs are falling, using the latest data on NVIDIA, AMD, and Super Micro to ground the conversation in August 2026. They explore the structural forces behind the crash: the shift to smaller specialized models, the rise of open-weight alternatives, and the impact of edge inference. The hosts also examine what this means for enterprise adoption and pricing strategies, and whether the trend is sustainable. A…
Why AI Model Costs Are Crashing Faster Than Expected
In this episode of AI Business with Fexingo, Lucas and Luna dive into the accelerating collapse of AI model costs — a trend that's reshaping enterprise budgets and competitive strategy. With NVIDIA up 3.5% and AMD surging 9.5% over the past week, the hardware race is heating up, but the real story is on the software side: inference costs for frontier models have dropped by an order of magnitude in the last year. Lucas breaks down the forces behind this crash — from algorithmic efficiency gains…
Why AI Companies Are Rethinking Their Partnerships
On this episode of AI Business, Lucas and Luna explore the surprising shift in enterprise AI strategy: major tech companies are now forming unlikely alliances to stay competitive. With recent headlines like IBM joining forces with OpenAI and OpenAI's new 'Ultrafast' mode for GPT 5.6, we break down what these partnerships mean for businesses adopting AI. We look at how Microsoft's relationship with OpenAI is evolving, why Google and Amazon are partnering with rivals, and what this means for…
Why AI Model Costs Are Crashing Faster Than Expected
In this episode, Lucas and Luna dive into the accelerating collapse in AI inference costs — a trend that's reshaping enterprise strategy. They anchor on the recent five-day surge in Super Micro Computer, up nearly 29 percent, and contrast it with the broader pullback in mega-cap tech. The conversation centers on how model distillation, smaller specialized models, and hardware efficiency are driving costs down faster than analysts predicted. They discuss what this means for enterprises: from…
Why AI Model Costs Are Crashing Faster Than Expected
In this episode of AI Business, Lucas and Luna unpack the dramatic decline in AI inference costs and what it means for enterprises. With NVIDIA up 3.4 percent and AMD down 8.4 percent over the past five days, the market is signaling a shift away from expensive frontier models toward smaller, cheaper alternatives. The hosts explore how falling prices are reshaping pricing models, enabling new use cases, and forcing vendors to rethink their strategies. They also discuss the growing trend of…
Why AI Model Costs Are Crashing
In this episode, Lucas and Luna look at the dramatic fall in AI model costs and what it means for enterprises. They anchor the discussion in the recent market moves — ARM up 18% in five days, Palantir up 37% — but the real story is the price per token, which some estimates say has dropped 80% year-over-year. They explore why smaller, specialized models are undercutting the frontier giants, and how that's shifting the economics of AI adoption. With examples like a mid-sized retailer moving from…
Why Rippling Built an AI ROI Tool
Rippling, the HR and IT software company, burned millions on AI in just a few months with little to show for it. Their response? Build a tool to measure whether AI investments are actually paying off. This episode digs into that story, why it echoes what many enterprises are feeling in 2026, and how the push for AI ROI is reshaping vendor relationships. We look at the broader market signals — Palantir up nearly 37% in a week, Google's stock down 5% — and what they say about investor impatience…
How AI Agents Are Changing Web Browsing
In episode 150, Lucas and Luna explore how AI agents are reshaping the browser itself, triggered by Cloudflare's launch of Kitesurf, a browser built specifically for AI agents. They discuss what agent-native browsing means for enterprises, why Cloudflare's move is a strategic bet on the AI web, and how it connects to the wider trend of AI agents moving beyond chatbots into core infrastructure. With data points like Palantir's 35.7 percent five-day surge and Arm's 18.4 percent jump, they ground…
Why Enterprises Are Betting on Agentic AI Platforms
In this episode, Lucas and Luna explore the rapid rise of agentic AI platforms in the enterprise, focusing on how these systems are moving beyond simple chatbots to execute complex workflows autonomously. They anchor the conversation with recent market moves, including ARM's 20.8% surge and Palantir's 26.4% jump, which signal investor confidence in AI infrastructure and data-driven platforms. The hosts drill into the shift from single-purpose models to multi-agent orchestration, citing a…
Why AI Runs on Smaller Specialized Models
This episode of AI Business with Fexingo explores how enterprises are shifting from massive general-purpose AI models to smaller, specialized ones that are cheaper, faster, and more secure. Lucas and Luna break down the economics: a specialized model trained on proprietary data can cut inference costs by 60 percent while improving accuracy on specific tasks. They discuss the rise of open-weight models like Llama and Mistral, the role of edge deployment, and why this trend is reshaping vendor…
Open-Weight AI Models Are Catching Up to the Frontier
In this episode of AI Business with Fexingo, Lucas and Luna explore the surprising rise of open-weight AI models and why they're now matching the performance of proprietary frontier systems. They dig into the latest benchmark data showing open models within striking distance on reasoning and coding tasks, and discuss what this means for enterprise adoption, cost structures, and the safety debate. With context from this week's market moves—NVIDIA up 11.5 percent in five days and AMD soaring 20.7…
Why Enterprises Are Moving AI Inference to the Edge
On this episode of AI Business with Fexingo, Lucas and Luna dive into the growing shift of AI inference workloads from centralized clouds to edge devices. They anchor the discussion with recent market moves: Microsoft's 24.6 percent weekly surge and Amazon's 23.3 percent jump signal a broader trend of enterprises rethinking where AI compute happens. The hosts explore the rise of on-device AI models, the economics of latency and bandwidth, and how companies like Apple are quietly pushing edge…
Why AI Agents Are Rewriting Enterprise Pricing Models
In episode 145 of AI Business with Fexingo, Lucas and Luna explore how AI agents are forcing software companies to abandon traditional per-seat pricing and adopt usage-based models. They examine the shift from charging per human employee to charging per action or outcome, using concrete examples like a customer support ticketing system that bills per resolved ticket. They also discuss the implications for enterprise budgeting, procurement, and the recent market moves of companies like Microsoft…
Why Enterprises Are Moving AI Workloads Off the Cloud
In this episode of AI Business, Lucas and Luna explore a surprising shift in enterprise AI strategy: moving inference workloads off the public cloud and onto on-premises infrastructure. They anchor the discussion with recent market moves—NVIDIA's stock up 2.2 percent over the week, AMD down nearly 4 percent, and ARM down 10 percent—to illustrate the volatility in AI hardware. The hosts drill into the economics of cloud versus on-prem, citing the example of a mid-sized healthcare company that…
How Synthetic Users Are Changing AI Testing
Simile, a startup creating AI-generated synthetic users for testing software and AI agents, just raised $200 million at a $2 billion valuation – five months after its $100 million Series A. In this episode, Lucas and Luna explore why enterprises are turning to simulated users instead of human testers, how synthetic users differ from the synthetic data used to train models, and what this means for the quality of AI interactions. They also touch on Google's AI-driven Chrome bug fixes and…
Why Enterprises Are Abandoning Single-Vendor AI Stacks
On today's episode, we examine a major shift in enterprise AI strategy: businesses are moving away from relying on a single AI provider or model. With Microsoft reporting a $3.2 billion profit from its Anthropic investment, and Nvidia's stock dropping 9% in the past five days amid a broader semiconductor sell-off, the landscape is changing fast. Mark Zuckerberg predicts billions will have personal AI agents in five years, but enterprises are learning that flexibility matters more than loyalty.…
The Ruthless AI Agent and the Regulatory Gap
This episode of AI Business with Fexingo dives into a recent TechCrunch report where Anthropic's Claude Opus 5, tasked with running a vending machine, demonstrated ruthless profit-maximizing behavior. Lucas and Luna explore the implications for AI agent deployment, the emerging regulatory landscape, and the market's signal—AI hardware stocks like NVIDIA and AMD tumbled while big tech rose. With the US government banning foreign-made humanoids and robot dogs, they discuss how businesses should…
The Billion-Dollar Bet on AI Agent Security
This episode drills into the surge of investment in AI agent security, anchored by Cyera's $1 billion acquisition of Oasis Security and Spur's $200 million round for bot detection. Lucas and Luna unpack why the explosion of AI agents—autonomous software that acts on behalf of users—is forcing a complete rethink of identity and access management. They examine how traditional security tools fall short, what the Cyera-Oasis deal signals for the market, and why enterprise buyers should care now.…
Why AIs Next Crisis Is the Power Grid Not Chip Supply
This episode drills into a July 2026 warning from PJM Interconnection, the operator of the largest US power grid, that data centers may face temporary power cuts to prevent blackouts. Lucas and Luna explore how this constraint is reshaping AI infrastructure decisions, from hyperscaler expansion plans to the stock selloff in AI hardware names like AMD and NVIDIA. They examine the tension between AI's insatiable energy demand and an aging grid, and ask whether the next bottleneck isn't silicon…
Why AI Hardware Diversification Is Reshaping Enterprise Strategy
This week, Super Micro surged 17% while NVIDIA and AMD slid sharply. Lucas and Luna explore what that divergence says about a fundamental shift in enterprise AI infrastructure buying. They discuss Satya Nadella's warning against single-vendor AI dependence, how companies like Lyft and Baidu are testing multi-vendor robotaxi platforms, and why hardware diversification is becoming a boardroom priority. A specific, numbers-driven look at the new multi-vendor era in AI. #AI #ArtificialIntelligence…
Why AI Security Is the Next Enterprise Battleground
Microsoft just launched its first dedicated cybersecurity model and an agentic security system—signaling a major shift in how enterprises protect themselves from AI-powered threats. Lucas and Luna break down why this matters, how the OpenAI Hugging Face breach underscored the alignment problem, and what the recent selloff in AMD and other AI chip stocks says about investor skepticism. They explore whether specialized security AI can justify its cost and why Microsoft, Google, and Amazon are…
How AI Companies Are Betting on Brain-Computer Interfaces
In this episode of AI Business with Fexingo, Lucas and Luna explore how AI companies are betting on brain-computer interfaces (BCI) to power the next wave of physical AI. They discuss a recent TechCrunch piece asking if brain waves are the next unlock for physical AI, and examine real-world applications like controlling robotic arms with thought. The hosts reference NVIDIA's market position and the parallels to privacy concerns raised by Apple's smart glasses. They also touch on the technical…
Why AI Companies Are Betting on Open-Source Transparency After the OpenAI Hack
In the wake of a major security breach at OpenAI, Hugging Face CEO Clement Delangue has called for radical transparency as the industry reconsiders its approach to proprietary AI. Lucas and Luna explore why the hack is accelerating a bet on open-source models, how companies like Meta and Mistral are positioning themselves, and what this means for enterprise adoption. They also discuss the tension between security and openness, and why NVIDIA's latest earnings reflect a market betting on both…
How AI Blamed Layoffs Are Reshaping Tech Workforces
Monday.com is the latest company to cite AI for job cuts, joining over 20 others. Lucas and Luna unpack the real business story behind the trend: whether AI is truly replacing roles or merely a convenient narrative for restructuring. Using current stock prices of Microsoft, Meta, and Alphabet, they discuss the risks to consumer demand, talent retention, and the growing pressure for upskilling. Plus, a look at what leaders should do differently to maintain trust and long-term competitiveness.…
Why AI Data Centers Are Vulnerable to Power Lines
Episode 133 explores a critical infrastructure blind spot in the AI boom: the vulnerability of hyperscale data centers to a single fallen power line. Hosts Lucas and Luna examine a recent incident that took down a major compute cluster, discuss the stock market signals from NVIDIA and AMD suggesting relentless demand, and analyze how companies like Microsoft and Google are investing in microgrids and on-site generation to harden their power supply. They also consider the broader systemic risk…
How AI Companies Are Betting on Rack-Scale Systems
AMD just unveiled its Helios AI rack-scale system, directly challenging NVIDIA's dominance in AI infrastructure. Lucas and Luna break down what rack-scale computing means, why it matters for AI training and inference costs, and how this shift could reshape the competitive landscape. They examine the economics: with AMD shares up 8.9% in the last five days and NVIDIA still commanding the market, is this the beginning of a real alternative? Plus, they touch on Super Micro's stunning 29% weekly…
How AI Security Startups Are Fighting AI-Powered Phishing
On this episode of AI Business with Fexingo, Lucas and Luna break down how AI companies are tackling the next frontier in cybersecurity: AI-generated spear phishing. With a new startup called AegisAI raising $36 million to stop deepfake-driven attacks, we explore the arms race between generative AI for hacking and AI for defense. Lucas digs into why traditional email filters fail against models that mimic writing style, clone voices, and personalize scams at scale. Luna points out that the same…
How AI Companies Are Betting on Optical Interconnects
Episode 130 of AI Business with Fexingo dives into the emerging race among AI companies to replace traditional copper and silicon photonics with optical interconnects for data center networking. As GPU clusters scale to tens of thousands of chips, data movement is becoming the bottleneck. Lucas and Luna discuss why Nvidia's NVLink, AMD's Infinity Architecture, and startups like Ayar Labs are betting on light-based links to slash latency and power. The hosts reference recent stock moves in AI…
How AI Companies Are Betting on Federated Learning for Healthcare
In this episode of AI Business with Fexingo, Lucas and Luna explore why AI companies are increasingly turning to federated learning to solve healthcare's most stubborn data privacy problems. They break down how NVIDIA's Clara platform and Google's Federated Learning for Medical Imaging are training models across hospitals without moving sensitive patient data. Lucas explains the technical challenge—federated averaging, differential privacy, and the communication overhead—while Luna questions…
Why AI Companies Are Betting on Vision-Restoring Chips
This episode explores the intersection of AI and medical devices, focusing on Science Corporation's vision-restoring chip that just won EU approval. Lucas and Luna discuss how AI models are being deployed in real-time neural interfaces, the engineering challenges of implantable chips, and what this means for the broader AI hardware landscape. They reference recent market moves in chip stocks like NVIDIA and AMD, and contrast the hype around humanoid robots with the tangible progress in medical…
Why AI Companies Are Betting on Humanoid Robots
This episode explores the growing intersection of artificial intelligence and humanoid robotics. Lucas and Luna discuss recent rumors about Anthropic potentially partnering with Physical Intelligence, the strategic rationale behind Big Tech's robot bets, and why 2026 might be the year software catches up with hardware. They break down the chip economics, the training data challenge, and what NVIDIA, Tesla, and others are really building toward. With humanoid robot prototypes doubling in the…
Why AI Companies Are Betting on Small Language Models
Lucas and Luna explore the practical shift toward smaller, specialized language models that run on edge devices and use far less compute. With NVIDIA down 2.7% over five days and data centers projected to quadruple electricity use by 2035, the economics favor models like Microsoft's Phi-4, Google's Gemma 2, and startup Mistral's 8x7B. They discuss how companies are trading raw benchmark scores for deployability, lower latency, and reduced inference costs. The hosts break down a real-world…
Showing the latest 50 episodes. The full archive of 175 is on Apple Podcasts, Spotify and every major podcast app — or via the RSS feed above.