00:00 Accountability at the Frontier
03:28 AI Tools and Workflows for Startups
08:09 Music, Creativity, and AI
09:53 The Evolution of Customer Onboarding with AI
16:52 Addressing the Needs of Small to Mid-Sized Businesses
17:54 The Future of AI: Ethics and Accountability
27:01 Navigating the Narrative Storm
30:02 The Risks of Autonomous Systems
31:41 Balancing Innovation and Regulation
33:17 Addressing the Cybersecurity Gap

Episode 23: AI for the Overlooked Customer

Our Guest: Justin Vandehey / Thread

Some customers never had a dedicated team to help them get software working. Justin Vandehey, co-founder of Thread, joins Justin Mannhardt and Kellan Danielson to discuss agents that fill that gap, and a broader question about who gets the benefits and protection AI can offer.

The Drip:

• How Vandehey uses AI for sales follow-up and shared intelligence across a distributed founding team.
• OpenClaw, AI music, and changing expectations about human and AI work.
• The hosts’ debate about frontier-model risk, accountability, and the gap between chat and agent experiences.

Inside The Bottle:

• Why customers keep explaining their goals after the sale is already closed.
• Proactive onboarding and activation versus reactive customer support.
• Thread’s work with Customer.io and its focus on helping small businesses start accepting payments.
• “There just weren’t people”: serving customers who never had dedicated implementation help.
• Years of useful work with today’s AI capabilities, even as frontier risks remain unsettled.
• Why smaller businesses and their suppliers face different questions about access to AI protection.

Related resource: Choosing where to start in your own business? AI Pathfinder offers a free personal Brief from Tiny Bottle AI’s co-founders after ten questions about your business.

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.

00:00 Introduction to AI Enablement
01:55 AI in Education and Quality of Work
03:35 The Role of AI in Problem Solving
04:09 Human Involvement in AI Processes
04:42 Navigating AI Recommendations in Consulting
07:20 The Balance of AI and Human Expertise
09:56 Evolving Use of AI in Professional Settings
13:51 Navigating the Non-Deterministic Nature of LLMs
15:55 Building Trust in AI: The User Experience
18:30 Understanding AI Integration Challenges
20:27 Establishing AI Strategy in Organizations
24:13 Avoiding the AI Hype: Practical Considerations
25:19 Fostering Adoption: Meeting Users Where They Are
32:10 Critical Thinking in the Age of AI

Episode 22: The AI Trust Bucket

If someone produces better work with AI, how do you know they're getting better at their job? Justin and Kellan talk about trust, accountability, and what it takes to help people use AI well. Along the way: a confident recommendation that needed another question, a camp agent named George, and knowing when ordinary process improvement is enough.

The Drip:

• The OpenAI research discussed this week: quality output, reasoning, and the question of whether people are building expertise.
• The Anthropic coding study and what people could explain after completing the task.

Inside The Bottle:

• A full trust bucket still needs boundaries.
• Why a promising demo needs to survive a typical month's work.
• Getting the leadership team aligned before choosing another tool.
• Meeting people where they are: three useful jobs for a camp agent.
• Keeping learning and accountability in the work, including the questions a manager will ask.

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.

00:05 The Impact of AI on Work Efficiency
04:25 Redefining Job Roles in the Age of AI
07:47 Creating Value Beyond Efficiency
15:52 The Importance of Strategy in AI Implementation
22:02 Governance Tools and Enablement
23:09 The Role of Strategy in Governance
24:19 Legal Ramifications and Governance Practices
26:46 AI Efficiency and Leadership Messaging
28:04 The Unanswerable Question of AI Benefits
31:11 Serving Society Through AI

Episode 21: Efficiency Isn't a Strategy

AI is creating real efficiency inside companies, but efficiency does not tell anyone what to do next. Justin and Kellan talk through the gap between saving time and creating value, why workforce decisions need a longer view, and how strategy gives governance and experimentation somewhere useful to aim.

The Drip:

• A BCG finding on frontline AI users saving roughly eight hours a week—often with little guidance for how to use the time
• Gartner reporting discussed in the episode: headcount reduction showed no clear correlation with better AI ROI
• Algorithmic pricing and the uncomfortable choices companies can make when optimization becomes the strategy

Inside The Bottle:

• Why AI-created efficiency is capacity, not automatically a better business result
• How automating entry-level work can weaken the future talent pipeline
• Why a compelling vision should come before policies, controls, and guardrails
• The two jobs of governance: mitigate risk and create clarity for good work
• The unresolved leadership question: who benefits from the eight hours AI gives back?

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.

00:00 Pete's Journey into AI and Renewables
05:33 The Role of AI in Sales and Personalization
08:07 Building Tools with AI: First Experiences
10:46 AI as a Facilitator in Decision Making
13:46 The Human Touch in AI Interactions
16:13 Trust and AI Adoption in Organizations
21:06 Building a Culture of Trust in AI Implementation
24:06 Connecting People to AI Tools Effectively
25:19 Preserving Trust and Authenticity in Sales
28:41 Designing Efficient Data Centers for AI
32:11 The Intersection of AI and Renewable Energy
34:29 Environmental Concerns and Technological Solutions

Episode 20: AI Is for Dummies?

Pete Halupka joins Justin and Kellan to talk about moving from reflexive AI skepticism to informed curiosity. They work through practical sales use cases, human quality gates, how companies discover an AI strategy, and the energy infrastructure beneath every prompt.

The Drip:

• AI activity is skyrocketing while public sentiment is getting worse
• Data-center demand is reshaping the conversation around the grid, solar, and battery assets
• Better model and infrastructure efficiency matter as much as adding more capacity

Inside The Bottle:

• The “AI is for dummies” banner that pushed Pete to form an opinion through use
• HubSpot enrichment and an ElevenLabs cold-call coach as practical sales tools
• Toyota's human quality gates as a model for responsible automation
• Why an AI strategy should be discovered inside existing business priorities
• Preserving trust, authenticity, and real relationships in AI-assisted sales
• Pete's view from renewable energy and the data-center buildout

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.

00:00 Embracing Change: The Journey to Baldness
02:31 The Freedom of Letting Go: Personal Growth and Self-Expression
05:08 AI in Everyday Life: The Balance of Automation and Personal Touch
07:47 Navigating AI: The Importance of Intentionality and Context
09:00 The Activity Trap: Understanding AI's Role in Workflows
15:23 Developing an AI Strategy for Companies
20:04 Personal AI Strategies and Individual Engagement
23:04 The Human Element in AI Interactions
26:14 Navigating AI's Role in the Workplace

Episode 19: The AI Activity Trap

Justin and Kellan start with Gemini moving into Chrome, then use a simple flight-booking example to ask the harder question: when does AI activity actually make a workflow better? They work through the gap between access and outcomes, why slower adoption can be smarter, and how strategy, ownership, and human judgment keep capable tools pointed at useful work.

The Drip:

• Gemini Spark can browse Chrome using a person's existing accounts and logged-in context.
• Distribution inside a tool people already use may matter more than another standalone AI destination.
• Being able to hand off a task such as booking a flight does not automatically make the experience better.

Inside The Bottle:

• Seats, prompts, tokens, and visible activity do not prove that a workflow or its outcome improved.
• Small, incremental deployments make it easier to discover missing context, rules, exceptions, and review checkpoints.
• A useful AI strategy clarifies what the company is trying to do, what it will not do, and who owns the result.
• The best automation boundary often appears where repetitive steps end and human judgment begins.
• The practical question is not only whether AI can do the work, but what becomes different when it does.

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.

00:00 Christine's Background and Role in AI
01:38 First Experiences with AI and Building Projects
04:40 The Joy of Making and Building with AI
06:23 Understanding User Needs in AI Development
07:25 The Balance of Speed and Quality in AI Design
12:24 The Role of Experience in AI Utilization
14:33 Mediocrity vs. Excellence in Design
16:08 Navigating AI in Creative Processes
18:09 Empowering Creatives with AI Tools
20:25 Emphasizing Human Creativity in AI
21:41 The Importance of Foundational Skills
23:44 Collaboration in the Age of AI

Episode 18: The Fizz with Special Guest Christine Sheller: Creative Work in the Age of AI

Justin and Kellan sit down with design leader and AI enablement consultant Christine Sheller to talk about what actually changes when creative teams can build at the speed of an idea. They get into taste, context, the frustrating last mile, and why more output does not automatically mean better work.

The Drip:

• Figma Make turns a workshop synthesis into a clickable prototype in about an hour.
• Image generation is getting better at the first draft, while refinement still takes work.
• As tools become widely available, creative teams need a real seat in the conversation about how they are used.

Inside The Bottle:

• AI can get an idea to 80% quickly; the last 20% is where judgment and craft show up.
• Design fundamentals, domain context, and experience still change the quality of the outcome.
• “AI slop” is often a brand-context problem, not simply a tool-use problem.
• Fast creation creates a new change-management problem: people cannot consume new work at the same speed.
• Collaboration and shared context matter more as everyone builds their own work products.

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.

00:05 Generational Perspectives on AI

02:26 Navigating AI in the Workplace

03:30 The Role of AI as a Thought Partner

06:24 The Dissonance of AI Expectations

08:38 The Future of AI and Human Creativity

18:37 The Value Exchange in AI Data Usage

20:09 Navigating AI in the Workplace

21:57 The Challenge of Institutional Knowledge

26:32 The Chaos of AI Outputs

29:59 Balancing Human and AI Contributions

36:34 From Ideas to Implementation in AI

39:32 Trust and Adoption in AI Solutions

Episode 17: Why Your AI Always Picks 46

What happens when everybody asks AI for advice and gets some version of the same answer? Justin and Kellan look at the growing trust gap around workplace AI, the explosion of artifacts and apps inside teams, and why adoption still depends on human judgment.

The Drip:

• Gallup's Gen Z AI-at-work numbers: excitement fell while anger climbed
• The gap between the workplace mandate to "figure out AI" and what people actually experience
• Why access to capable tools changes how believable the broader AI narrative feels

Inside The Bottle:

• Why a model asked for a number between zero and one hundred keeps landing on 46
• The "AI slop loop" and what happens when training data starts feeding on AI-generated work
• How personal context files and individualized agent behavior create institutional-knowledge chaos
• Why more apps, documents, and tokens consumed do not equal useful adoption
• A crawl-walk-run path from chat to harness to API, with humans kept in the places that require judgment

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.

00:05 Cognitive Offloading and AI's Impact on Thinking

02:38 The Ethics of AI in Decision Making

05:39 Accountability in AI-Driven Decisions

06:02 Navigating AI Limitations and Human Judgment

10:49 The Role of AI in Content Creation

14:10 Authenticity and Value in AI-Generated Content

17:40 Exploring AI's Decision-Making Capabilities

18:48 Daily AI Routines and Their Impact

20:11 The Importance of Human Decisions in AI

21:45 Understanding AI's Contextual Limitations

22:48 Navigating AI's Memory and Contextual Challenges

22:48 The Role of AI in Accountability and Truth

24:26 Cultural Implications of AI in the Workplace

25:11 Ethics and Human Touch in AI Interactions

27:24 AI as a Tool for Knowledge and Understanding

28:51 Maintaining Agency in an AI-Driven World

Episode 16: Eight Human Decisions a Day

Are we handing too much of our thinking to AI? Justin and Kellan take on the week's big question, from Meta's lawsuit over AI-assisted layoff rankings to the boundaries they draw in their own daily work, and get specific about what's worth handing off and what should stay stubbornly human.

The Drip:

• Meta sued by former employees over AI-assisted layoff decisions, and why the fight is about accountability, not capability
• The "are we offloading our thinking?" essays making the rounds, and where AI-first tips into AI-only
• Companies using AI avatars for first-round phone screens (hard pass)

Inside The Bottle:

• The boundary question: what should AI think about and decide, and what should you?
• Why "I can't believe Claude did that" is really a story about what you fed it
• Justin's grounding rule: confirm current state first, then decide what happens next
• Kellan's daily assistant that counts human vs. autonomous decisions, and the eight-a-day rule
• Constraining the blast radius: how bad is it if this thing makes a mistake?
• Keeping your voice: the 15-minute writing rule and AI cleanup that doesn't change your words

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.

00:05 The Rise of AI Coworkers
02:16 Navigating Confusion in AI Tools
04:06 The Challenge of User Experience
05:16 The Race for Market Share
07:47 Understanding AI's Agentic Capabilities
10:10 The Complexity of AI Interfaces
10:11 Rethinking AI Nomenclature
11:57 The Future of AI Collaboration
15:27 Navigating AI Workspaces
17:01 The Challenge of AI Integration
20:06 Product Maturity and User Experience
22:04 Understanding AI Tools and Their Applications
24:51 The Future of AI Models and Software
28:06 Simplifying AI Workflows

Episode 15: The Coworker Craze

OpenAI, Meta, Google, and Anthropic all shipped their version of an "AI coworker" in the same week. Justin and Kellan sort out what's actually new, what's just a new noun, and how they actually split their own work across chat, code, and everything in between.

The Drip:

• OpenAI releases ChatGPT Work — and the app rollout that has everyone confused
• Microsoft standardizes Copilot on GPT 5.6
• Meta ships Spark for agent orchestration; Gemini launches new features
• Why "agentic AI for business people" still has no obvious home

Inside The Bottle:

• Kellan's hot take: get rid of the tabs — one surface, let the model route the work
• Cowork and Claude Code are fundamentally the same thing; one is just more gated
• Justin's rule for picking a tool: does this thing have a runtime, or does it need shared context?
• Chat, plan, build — the mental model that actually holds up
• The subscription hack: same capabilities, roughly 2x the tokens
• The two things worth staying on top of: model progress, and the software around the models
• "The model is the person. The tool is how you choose to talk to them."

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.

00:00 Introduction and Mail Surprises

01:16 Fable Five and AI News

03:07 Comparing Fable and Opus

04:20 The Evolution of AI Intelligence

04:48 Practical Applications and Experiences

06:29 Managing AI Memories and Workflows

08:48 Optimizing AI Usage and Budgeting

11:00 Understanding AI Model Choices and Effort Levels

11:20 Introduction to Pathfinder

13:57 The Inspiration Behind Pathfinder

14:21 Identifying Focus and Alignment Problems

15:22 Demonstrating Pathfinder's Features

16:39 The Brainstorming Process

17:25 Focusing Ideas for Implementation

18:49 Evaluating Ideas and Avoiding Pitfalls

22:17 Building with Purpose: The Importance of Focus

24:17 Creating a Structured Build Plan

26:34 Establishing a Roadmap for Success

30:01 Final Thoughts and Team Collaboration

Episode 14: You Don't Have an AI Problem

Pathfinder is real, it's live, and Justin and Kellan finally take you on the full tour: why they built it, how the four-step method works, and what it looks like to go from "we should do something with AI" to a project you're actually building. But first: Fable 5 leaves the Anthropic subscription, and what that means for how practitioners should stack their models.

The Drip:

• Fable 5 leaves Anthropic subscriptions: the July 7 deadline got a last-minute extension to July 12
• Kellan's setup: downgrade the plan, add usage credits, and run Fable as the planner and reviewer while cheaper models implement
• Memories have gotten out of control, and it's time to ask Claude for a full memory audit
• Effort levels: why medium or high beats extra-high and ultracode for most real work

Inside The Bottle:

• Pathfinder is live and free for individuals at pathfinder.tinybottleai.com
• The backstory: you don't have an AI problem, you probably have a focus problem
• The ideas brainstorm: the AI interviews you. Fifteen minutes gets you six to ten real ideas, no blank page
• Focusing an idea into a Brief: the real problem, what solving it means, and the right solution type (automation, assistant, or agentic)
• The Build Plan: the smallest first version, handed off to Claude Code, Codex, or ChatGPT via copy, MCP, or CLI
• Roadmaps and checkpoints: the commitment system that keeps you moving after the excitement wears off
• The team plan: idea synthesis and prioritization across your whole team, with Justin and Kellan involved

Want to find your best AI project? Pathfinder is free for individuals. Sign up at pathfinder.tinybottleai.com and see what it turns up. If you need a hand, support goes straight to Justin and Kellan.

New episodes every Wednesday. Subscribe to The Weekly Fizz newsletter at tinybottleai.com/the-fizz.