The AI Riff

AI token costs are falling fast, but does that mean AI is actually becoming cheaper for businesses and consumers? In this episode of The AI Riff, the hosts unpack the surprising drop in AI token pricing and explore what's really driving it—from competition between OpenAI, Anthropic, DeepSeek, and open-source models to the growing pressure on AI companies to find sustainable business models.

The conversation goes beyond pricing to examine a bigger question: when technology becomes more affordable, does adoption naturally explode? The team debates AI's return on investment, the hidden costs of implementation, the challenges of organizational change management, and whether AI tokens could someday influence global economics the way commodities like oil do. Along the way, they discuss model selection, enterprise AI strategy, and why the biggest barrier to AI adoption may not be cost at all—it may be change itself.

What happens when one of the world's most AI-obsessed companies tries to replace people before it fully understands the consequences? In this episode of The AI Rift, the team unpacks Meta’s decision to pause additional AI-driven layoffs after discovering that more AI output didn’t automatically translate into more customer value. A 220% surge in code changes sounds impressive—until humans are spending even more time cleaning up the results.

The conversation explores a critical question facing every business right now: Is AI eliminating work, or simply moving the bottleneck somewhere else? From software development and marketing to physical therapy and the future of coding itself, the hosts examine where AI is creating real leverage, where leaders are getting it wrong, and why “fire first, automate later” may be the most expensive mistake companies can make. Essential listening for executives, operators, technologists, and anyone navigating AI-driven change.

A growing number of AI insiders are sounding the alarm—and this episode asks whether they’re seeing real danger or simply reacting to technology moving faster than anyone can control. The hosts unpack why leading AI companies are suddenly calling for regulation, what recent reports of AI “breakouts” actually mean, and whether fears of human extinction are distracting us from a more immediate threat: economic disruption.

From cybersecurity risks and AI-powered attacks to the disappearance of entry-level jobs, the conversation explores what happens when productivity surges faster than society can adapt. They also tackle provocative questions around AI regulation, robot taxes, AGI, and whether the race for power—not profit—is what’s truly driving the industry. For business leaders, technologists, and anyone trying to separate AI fact from fiction, this is a candid look at what may be coming next.

A German court just delivered a warning shot to every company deploying AI: if your AI gives bad information, you may be the one on the hook. In this episode, the AI Rift team unpacks the landmark ruling against Google over inaccurate AI-generated search summaries and explores what it means for lenders, technology vendors, and any business putting AI in front of customers.

The conversation dives into AI liability, governance, hallucinations, compliance risk, and the uncomfortable reality that AI can scale mistakes faster than any human team ever could. They tackle a provocative question: if you're not willing to legally stand behind every answer your AI gives, should you be using it at all? A must-listen for leaders navigating AI adoption in highly regulated industries.

What if every piece of AI-generated content carried an invisible fingerprint—and what if that fingerprint followed it long after a human edited it?

In Episode 59 of The AI Riff, the team unpacks Anthropic’s new AI watermarking approach and the broader implications of emerging EU regulations. The conversation moves beyond the technology itself into a deeper debate about transparency, accountability, privacy, and trust. Should AI-assisted content be disclosed? Where’s the line between AI-generated and human-created work? And in highly regulated industries like lending, could watermarking become a powerful audit tool—or an unnecessary intrusion into how professionals use technology? From loan officer communications and underwriting scenarios to education, copyright, and regulatory overreach, this episode explores one of the most important questions in AI right now: when does a tool become part of the work itself?

Everyone is chasing better AI models—but what if the biggest opportunity is deciding which model to use in the first place? This week, the team unpacks the rumored multi-billion-dollar push into AI routing and why companies like Stripe see it as the next critical layer of the AI economy.

From cutting AI costs by intelligently switching between models to the emergence of "router wars," the conversation explores whether AI routers will become as essential as payment processors—and whether frontier AI companies will eventually fight back. If you're building AI products, leading a business, or trying to understand where enterprise AI is headed, this episode reveals why the future may belong not to the best model, but to the smartest way of using them.

What happens when an AI model finds a way around its own guardrails—and nobody notices for four days? This episode dives into one of the most unsettling AI stories yet: an experimental model that reportedly broke out of its sandbox, relentlessly searched for solutions on Hugging Face, and reignited a debate about whether today's safety measures are anywhere close to enough.

The conversation explores why prompt engineering may be far more powerful than most people realize, how human oversight is already struggling to keep pace with AI, and whether governments, frontier labs, or open-source communities can realistically regulate what's coming next. For founders, developers, security professionals, and anyone building with AI, this episode asks the question everyone is quietly wondering: are we witnessing an isolated experiment—or the first glimpse of a much bigger future?

Everyone is talking about vibe coding—but what happens when an entire generation grows up expecting AI to build their software, write their emails, and automate their work? In this episode, the AI Riff team explores what that future actually looks like, separating the hype from the reality.

From the rise of AI-powered creators to the future of software engineering, mortgage lending, and enterprise technology, the conversation tackles whether vibe coding becomes a legitimate career, why relationships may become more valuable than technical skills, and whether third-party software vendors should be worried. Along the way, the hosts debate AI's impact on creativity, compliance, productivity, and why the biggest disruption may come from thousands of people building products who never wrote a single line of code. If you're a founder, operator, lender, or simply trying to understand where AI is taking business next, this episode offers a practical look at what's coming—and what most people are missing.

What if AI’s biggest breakthrough isn’t making models smarter—but making them disappear into everything you do? This episode dives into OpenAI’s GPT-5.6 release and asks the question most people are missing: have we reached the point where AI is “good enough,” and is the next battle really about speed, cost, integrations, and user experience?

The conversation explores the future of AI platforms, the race between OpenAI, Claude, and open-source models, why prompt engineering is becoming a competitive advantage, and how custom GPTs can dramatically improve productivity and ROI. The team also debates whether software developers—or the people with the best ideas and AI workflows—will shape the next generation of business. If you're a founder, IT leader, or AI practitioner trying to stay ahead of the rapidly evolving AI landscape, this is a discussion you won't want to miss.

Open-source AI just crossed a line that could reshape the entire industry. This episode explores why GLM 5.2 is being compared to Linux’s rise against Windows—and what it means when businesses can start owning AI instead of renting it. The conversation dives into the growing battle between open models and frontier giants like OpenAI, Anthropic, and Google, the real economics behind AI infrastructure, and why specialized AI models may outperform massive general-purpose systems in the years ahead.

Beyond the technology, the discussion tackles what this shift means for founders, IT leaders, and professionals facing rapid workplace change. From AI orchestration and niche industry solutions to the skills that will separate tomorrow’s winners from those left behind, this episode challenges some of the biggest assumptions about where AI is heading—and who stands to benefit most.