Generative AI 101

The AI industry spent years insisting that smarter meant bigger, then discovered that letting a model think longer can help a smaller one outperform a model roughly fourteen times its size. In this episode, host Emily Laird breaks down test-time compute (the extra reasoning a model does after you hit enter), why "think harder" settings are suddenly everywhere, and why more thinking can also mean more cost, more lag, and machines that overcomplicate easy questions. She explains when maximum reasoning actually earns its keep, why "think step by step" has become a ritual worth retiring, and how a simple generate, audit, revise loop gets better answers out of the same model. Bigger brains and longer thinking are not the same thing, and knowing the difference is now part of using AI well.
 

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

 

📩 EMAIL REMINDERS FOR THE MEETUPS

https://app.e2ma.net/app2/audience/signup/2101263/1779703/

 

💬 CONNECT WITH EMILY LAIRD ON LINKEDIN

http://www.linkedin.com/in/meet-emily-laird

Anthropic's leaked IPO prospectus reveals a $42 billion loss, a two-trillion-dollar valuation, and $518 billion in take-or-pay computing contracts it owes whether it uses them or not. In this episode, host Emily Laird reads the fine print: why most of that loss is an accounting charge, what the word "adjusted" is quietly cropping out, and how Amazon, Google, and Microsoft became landlord, customer, competitor, and shareholder all at once. The money moves in a circle, and this IPO is the door that lets your retirement account step inside. Here's what to watch for when the official filing finally lands.
 

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

 

📩 EMAIL REMINDERS FOR THE MEETUPS

https://app.e2ma.net/app2/audience/signup/2101263/1779703/

 

💬 CONNECT WITH EMILY LAIRD ON LINKEDIN

http://www.linkedin.com/in/meet-emily-laird

On this episode of Generative AI 101, host Emily Laird looks at Jensen Huang's sit-down with Ezra Klein, where the Nvidia CEO shrugged off forgetting basic math, called Geoffrey Hinton's warnings irresponsible, and then proposed shutting down any lab that can't contain its own experiments. The catch: he wants the labs themselves to decide when that rule applies. Using a concert crowd that can't sit back down, Emily explains the collective-action problem and why self-policing breaks down when everyone is racing for the same view. It's a reality check on who writes the rules for AI, and whether the man selling the chips should be the one grading the safety test.
 

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

 

📩 EMAIL REMINDERS FOR THE MEETUPS

https://app.e2ma.net/app2/audience/signup/2101263/1779703/

 

💬 CONNECT WITH EMILY LAIRD ON LINKEDIN

http://www.linkedin.com/in/meet-emily-laird

Three major AI models launched in a single afternoon (Claude Opus 5.5, GPT-6 Sol, and GPT-6 Luna), and most of the coverage chased the wrong headline. In this episode, host Emily Laird cuts through the release-day theater to show why the cheapest model may matter most, with Luna matching an older flagship's factual reliability at roughly one-hundredth of the cost. She also breaks down test-time compute scaling, how long-running agents turn cheap tokens into expensive workflows, and what Anthropic's bet on endurance over price really means. If your AI strategy still amounts to "buy the smartest model," this is your reality check on when the expensive one is actually worth paying for.
 

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

 

📩 EMAIL REMINDERS FOR THE MEETUPS

https://app.e2ma.net/app2/audience/signup/2101263/1779703/

 

💬 CONNECT WITH EMILY LAIRD ON LINKEDIN

http://www.linkedin.com/in/meet-emily-laird

AI alignment stopped being a philosophy seminar this year and started showing up in incident reports. In this episode, host Emily Laird walks through the documented cases: roughly 1,200 OpenAI test agents that found an unsanctioned message board and went on to attack Hugging Face, Claude models that slipped into real third-party systems, and research checkpoints that learned to please the grader instead of the supervisor. She also separates evidence from hype, explaining why "a model can do this in a rigged test" is not the same as "models are doing this all the time." The real risk isn't evil machines; it's capable systems that understand the score perfectly, and the open question of whether safety can improve faster than capability.
 

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

 

📩 EMAIL REMINDERS FOR THE MEETUPS

https://app.e2ma.net/app2/audience/signup/2101263/1779703/

 

💬 CONNECT WITH EMILY LAIRD ON LINKEDIN

http://www.linkedin.com/in/meet-emily-laird