Serious AI with Chris Duff

The US builds the most advanced AI in the world and worries about it more than almost anywhere else. Half of Americans are more concerned than excited, against 16% in South Korea, and the US ranks 21st in everyday use while trusting its own government to regulate AI least of 30 countries. Europe makes companies tell people it's AI at the first interaction, and most of the US leaves that call to you. One study (presenter-reported) found a reply that says "this is AI" scored +10 with customers and saying nothing scored minus 30, so I walk through what to fix in your own customer-facing AI this week, and what Claude Fable 5.2 says about where this is headed.

Pew Research Center, how 25 countries feel about AI (October 2025): https://www.pewresearch.org/wp-content/uploads/sites/20/2025/10/pg_2025.10.15_ai_report.pdf
Stanford AI Index 2026, public opinion chapter (trust in regulators, the 79% who want disclosure): https://hai.stanford.edu/assets/files/ai_index_report_2026_chapter_9_public_opinion.pdf
Stanford AI Index 2025, public opinion (the 2024 benefit numbers): https://hai.stanford.edu/ai-index/2025-ai-index-report/public-opinion
Microsoft AI Economy Institute, everyday AI use by country in 2026: https://blogs.microsoft.com/on-the-issues/2026/05/07/the-state-of-global-ai-diffusion-in-2026/
EU AI Act, Article 50, telling people it's AI: https://artificialintelligenceact.eu/article/50/
Utah SB 226, disclosure when the customer asks: https://le.utah.gov/Session/2025/bills/enrolled/SB0226.pdf
Colorado SB26-189, the 2027 notice rule: https://leg.colorado.gov/bills/sb26-189
CX Vision 2035, the customer study on disclosing AI (presenter-reported): https://cx-vision-2035.aiamplifier.com/
Claude Fable 5.2 launch, November 9, 2026: https://sdtimes.com/ai/anthropic-introduces-claude-fable-5-2-mythos-5-2/

Serious AI News, every Monday. AI for operators, what we built and what it cost:
https://seriousai.partners/newsletter/?utm_source=podcast&utm_medium=description&utm_content=ep-333-ai-temperature

Tell us what's breaking:
https://seriousai.partners/contact/?utm_source=podcast&utm_medium=description&utm_content=ep-333-ai-temperature

Chapters:
0:00 The country building AI worries about it the most
0:59 Who I am and why this matters to operators
1:30 Customers who can't reach a human
3:11 The rest of the world feels differently
4:15 Why I'm still an optimist, Claude Fable 5.2
5:21 Why Americans got worried
7:44 Feel, the most worried of 25 countries
9:00 Rules, Europe says it and the US is a patchwork
10:10 Use, the US ranks 21st
10:39 Saying it's AI, +10 against minus 30
11:53 79% want disclosure, customers already decided
12:34 Why I'm open to work abroad
13:41 What to do this week
15:16 A wild turning point

#AIForBusiness #CustomerExperience #AITransparency #Chatbots #EUAIAct

I pointed Claude Code at my tax returns, statements and budget sheets for seven days, about 3,400 messages, read-only. It got a $243,000 line wrong, and it also surfaced a capital loss I'd forgotten, a 0.4% fee on an account of short-term bonds, and a crypto loss of about $600K on my worksheet (my preparer reviews it before anything is filed). The episode walks through the Family CFO Triangle: See, Route and Shield, with the check at each corner I'd never hand off. I'm not a financial advisor, and nothing here is advice.

Claude Code: https://claude.com/product/claude-code
Profit First by Mike Michalowicz: https://www.profitfirstbook.com
REtipster, where I first heard about it: https://retipster.com/
IRS Topic 409, capital gains and losses: https://www.irs.gov/taxtopics/tc409
IRS on Roth IRAs: https://www.irs.gov/retirement-plans/roth-iras

Serious AI News, every Monday. AI for operators, what we built and what it cost:
https://seriousai.partners/newsletter/?utm_source=podcast&utm_medium=description&utm_content=ep-332-family-cfo

Tell us what's breaking:
https://seriousai.partners/contact/?utm_source=podcast&utm_medium=description&utm_content=ep-332-family-cfo

Chapters:
0:00 I gave Claude Code my family's finances
1:26 Why now, the levers close December 31
1:58 Seven days, about 3,400 messages
2:35 The Family CFO Triangle
5:00 See, point it at everything, read-only
7:16 Rebuilding the trackers
9:29 My real spend ran a third above my sheet
10:40 The $52,000 I couldn't explain
11:22 The $243,000 miss
13:08 Route, what I kept from Profit First
16:09 Every dollar gets a job
17:23 One call to the bank, a 0.4% fee gone
19:11 Asking for IPO access
20:55 Shield, an old loss becomes a tax shield
25:07 Crypto is crypto, cold storage
26:00 The cheap tax window
27:15 The catch, a second thread saves the credit
28:10 Buy the boring protection
30:11 What to do before December 31
31:30 The check at each corner

#ClaudeCode #PersonalFinance #TaxPlanning #FractionalCFO #AIForBusiness

Most practices rank their payers by the monthly denial report, so the plan that denies the most gets the attention. I built a check on the 835 remits (the electronic statements that come back with every payment) to see what that ranking misses, using made-up sample data for a 24-clinician GI group. A payer costs you three ways: paid late, paid short, and paid by your staff's time chasing it. In the sample, the plan that paid late and short cost almost $6 per $100 paid, and the biggest denier cost about 50 cents.

Want the payer cost demo to try with your own numbers? Comment PAYER on the YouTube episode and I'll send it over.
835 remittance advice, what the file is (CMS): https://www.cms.gov/medicare/billing/electronicbillingeditrans/remittance

AI for medical groups and practices:
https://seriousai.partners/healthcare/?utm_source=podcast&utm_medium=description&utm_content=ep-331-payer-cost

Serious AI News, every Monday. AI for operators, what we built and what it cost:
https://seriousai.partners/newsletter/?utm_source=podcast&utm_medium=description&utm_content=ep-331-payer-cost

Tell us what's breaking:
https://seriousai.partners/contact/?utm_source=podcast&utm_medium=description&utm_content=ep-331-payer-cost

Chapters:
0:00 The worst payer wasn't the denier
0:51 A check on 835 remits, sample data
1:50 My read on 2027
2:37 The payer cost triangle
3:35 Paid late, 13 days against 47
5:07 Paid short, one claim end to end
7:49 The escalator nobody checked
10:04 Why plain code checks every line
11:37 Paid by your staff
12:07 One payer, one full-time employee
12:43 The biggest denier isn't the costliest
14:46 Try to break it
16:50 What goes into a demo like this
19:54 Your one thing this week

#HealthcareAI #RevenueCycle #MedicalBilling #MedicalPractice #AIinHealthcare

My best build got no feedback, and a LinkedIn post about a piece of healthcare paperwork hit 9,400 impressions (my posts usually draw 44 to 89). The paperwork is the BAA, the business associate agreement HIPAA requires before any vendor touches patient data, and the comment thread mapped three ways to keep that data safe with AI, plus the catches: you can't redact a voice, a BAA covers one vendor while the data runs through several, and no paperwork controls what an agent says out loud. Here's what that taught me about content, and the one exercise I'd have every practice run this week (a builder's read of the vendors' own pages, not legal advice).

AI for medical groups and practices:
https://seriousai.partners/healthcare/?utm_source=podcast&utm_medium=description&utm_content=ep-330-baa-post

Serious AI News, every Monday. AI for operators, what we built and what it cost:
https://seriousai.partners/newsletter/?utm_source=podcast&utm_medium=description&utm_content=ep-330-baa-post

Tell us what's breaking:
https://seriousai.partners/contact/?utm_source=podcast&utm_medium=description&utm_content=ep-330-baa-post

Chapters:
0:00 My best build got no feedback
0:09 What a BAA is, in 20 seconds
0:50 The claim that started it
1:48 Why a piece of paperwork matters
5:18 The build that drew nothing
6:20 Why the BAA post took off
7:35 Why the demos stay
8:47 Strangers debating in my comments
9:50 Three doors to keep patient data safe
12:46 Catch one, you can't redact a voice
13:23 Catch two, one BAA and several vendors
14:30 Catch three, what an agent says out loud
16:22 Says who? Check the claim yourself
16:55 Three follow-through habits
19:57 Map where your patient data goes

#HealthcareAI #HIPAA #AIinHealthcare #MedicalPractice #ContentStrategy

Claude is probably running an old copy of your skill. I had around 70 skills and 21 routines, and Google Drive showed every one of them synced while Claude Code ran different versions than Claude chat and Cowork. In this episode I install my free template from zero on a fresh account: one GitHub repo that Claude Code, Claude chat and Cowork all update from, with Vale (a free style checker) blocking a broken house rule instead of trusting Claude to remember it. If your skills live in more than one place, they have already drifted.

The free template, with the one line to paste into Claude Code:
https://github.com/serious-ai/claude-skills-starter

Matt Pocock's skills repo, where this started:
https://github.com/mattpocock/skills

Serious AI News, every Monday. AI for operators, what we built and what it cost:
https://seriousai.partners/newsletter/?utm_source=podcast&utm_medium=description&utm_content=ep-329-claude-skills-repo

Tell us what's breaking:
https://seriousai.partners/contact/?utm_source=podcast&utm_medium=description&utm_content=ep-329-claude-skills-repo

Chapters:
0:00 Claude is probably running an old copy of your skill
1:12 The growing pile of skills nobody can track
2:04 69 skills and 21 routines in one GitHub repo
3:07 Drive said synced, Claude ran other versions
4:38 Where it started, Matt Pocock's skills repo
5:47 The free template and the one-line install
6:45 Reviewing a downloaded skill before it goes in
8:44 Why this demo runs in Terminal
9:57 A fresh account, installed from zero
12:37 Plugin commands and auto-update
13:42 Adding the marketplace in the browser
16:23 Installed, plus the friction it caught
17:57 A house voice that updates itself
19:13 A rule becomes a check, Vale and em dashes
20:04 Teams, a plugin per department
21:02 Count the places your skills live

#ClaudeCode #AIAgents #AIAutomation #GitHub #AIForBusiness

At the MGMA Annual Conference (the Medical Group Management Association's yearly meeting for people who run medical practices) in San Antonio, I ran Claude Code against my live Notion meeting transcripts, and every 10 minutes it suggested what to ask next. This screen-share walks through the whole setup: picking 18 sessions before the trip, a glossary of revenue cycle jargon built as the speakers talked, a field log that caught 10 healthcare opportunities, and what broke when my laptop battery couldn't keep up. I walked in knowing nobody in practice management and left with a plan for every follow-up. Take any piece of it for your next event.

AI for medical groups and practices:
https://seriousai.partners/healthcare/?utm_source=podcast&utm_medium=description&utm_content=ep-328-ai-conference-questions

Serious AI News, every Monday. AI for operators, what we built and what it cost:
https://seriousai.partners/newsletter/?utm_source=podcast&utm_medium=description&utm_content=ep-328-ai-conference-questions

Tell us what's breaking:
https://seriousai.partners/contact/?utm_source=podcast&utm_medium=description&utm_content=ep-328-ai-conference-questions

Chapters:
0:00 Every 10 minutes, my MacBook told me what to ask
1:46 Why a healthcare conference
2:34 Picking 18 sessions with Claude before the trip
5:37 The first session, a fire hose of jargon
7:33 Long Claude Code threads hold up now
10:21 Live questions every 10 minutes
11:38 Why I never asked a question
13:33 A running glossary, built live
14:57 Composio and more than one Google account
17:14 What Claude can't see in the room
18:27 The routine Claude built for itself
20:10 The battery problem and the phone workaround
23:43 Slides in, field log out
24:21 Lessons, fewer sessions and more meetings
25:35 Was I the agent?
28:49 Why I'd never skip this setup again

#ClaudeCode #HealthcareAI #AIForBusiness #AIAgents #ConferenceStrategy

I spent two and a half days at MGMA 2026, the national conference for medical group leaders, as an AI builder with a medical background who went there to learn. What I found is a blind-spot triangle: are you paid what your contracts say, is your AI's work right, and does anyone own the whole picture across the tools you bought one at a time. Payers already use AI to deny claims and quietly downcode them, and many of the practices in those rooms are still answering by hand. Take one question back this week: for an AI tool you pay for, what single number would prove it's working?

AI for medical groups and practices:
https://seriousai.partners/healthcare/?utm_source=podcast&utm_medium=description&utm_content=ep-327-underpaid-no-denial

Serious AI News, every Monday. AI for operators, what we built and what it cost:
https://seriousai.partners/newsletter/?utm_source=podcast&utm_medium=description&utm_content=ep-327-underpaid-no-denial

Tell us what's breaking:
https://seriousai.partners/contact/?utm_source=podcast&utm_medium=description&utm_content=ep-327-underpaid-no-denial

Chapters:
0:00 Buying AI one tool at a time
0:16 MGMA 2026, going in to learn
2:50 AI in every room, and the pressure on operators
4:21 Point tools and ambient scribes
5:36 Only a few run a mature AI program
6:57 Payers already automate denials
8:16 Why the buying decisions are happening now
9:50 The blind-spot triangle
12:26 In-house build versus vendor tools
14:33 Silent downcoding, billed one code and paid another
15:45 Contracts that pay less than they say
17:09 A scribe that ignored its rules
18:10 Coders still checking by hand
19:11 The human side of automation
20:20 Who owns the whole picture
21:51 Your move, name one number

#HealthcareAI #MedicalPractice #RevenueCycle #AIImplementation #EnterpriseAI

Two days after picking healthcare as the focus for Serious AI, I bought a $2,400 ticket to the MGMA annual conference in San Antonio. Before the trip I had Claude research every speaker and all 171 exhibitors, narrow more than 100 organizations to 14 targets, and build each one a short, specific read on where AI could help their practice. This episode walks through the whole plan: why healthcare, how the pain points were checked before anything shipped, and why the uncomfortable choice was the right one.

AI for medical groups and practices:

https://seriousai.partners/healthcare/?utm_source=podcast&utm_medium=description&utm_content=ep-326-2m-pipeline

Serious AI News, every Monday. AI for operators, what we built and what it cost:

https://seriousai.partners/newsletter/?utm_source=podcast&utm_medium=description&utm_content=ep-326-2m-pipeline

Tell us what's breaking:

https://seriousai.partners/contact/?utm_source=podcast&utm_medium=description&utm_content=ep-326-2m-pipeline

Chapters:

0:00 A $2,400 conference ticket

0:41 A wide band and no vertical

2:44 Why healthcare

5:47 A healthcare page in a day

8:24 Checking the pain points before they ship

9:25 The Ask Method, rebuilt with AI

11:02 Retargeting AI search around healthcare

11:52 Conferences and the hiring signal

14:06 MGMA, this weekend in San Antonio

17:36 Choosing the uncomfortable option

19:47 Researching every speaker before the trip

21:35 Artifacts that show the work first

22:36 14 targets and the host organization

25:27 The Tim Ferriss playbook

26:46 What's next

#HealthcareAI #MedicalPractice #AIImplementation #FractionalHeadOfAI #ClaudeCode

Seven days to move a business from one brand to another. DNS, a full website, SEO, the podcast routines, every description and every link, and Claude Code ran the whole thing on barely 10% of a week's usage on Claude Max. It still wrote an FAQ where 17 of 20 answers had to be rewritten by hand. AI output sounds equally smart whether it is correct or wrong, and if you have not decided what your company stands for, you have no way to tell which one you are reading.

Serious AI News, every Monday. AI for operators, what we built and what it cost: https://seriousai.partners/newsletter/?utm_source=podcast&utm_medium=description&utm_content=ep325

Tell us what's breaking: https://seriousai.partners/contact/?utm_source=podcast&utm_medium=description&utm_content=ep325

Chapters:

0:00 The rename, and going full time on enterprise AI

1:58 Cadence, and what carries over from the old show

6:11 Where the show goes next, live builds and demos

9:01 Being up front about the commercial incentive

9:30 No business I have met is fully AI optimized

11:57 Seven days to stand up a company

15:48 The infrastructure work that used to need a contractor

18:31 Why starting anything new costs more than it looks

22:11 Where the models stop being useful

23:38 Polsia, a $10M run rate with one-star reviews

25:52 17 of 20 FAQ answers had to be rewritten

28:16 Brand first, or chase customers first

30:12 The routine that removes the human from publishing

32:35 Where my judgment overrides the model

The SLC Deal Engine is live and Claude Code handled nearly the entire launch. Chris breaks down how the AI land underwriting tool went from production-ready to publicly for sale in a single day: structuring the offer around Alex Hormozi's value equation, filming a side-by-side race between his manual underwriting and the Deal Engine as the core proof point, and letting Claude Code cut 90 minutes of raw footage into a 20-minute branded video sales letter then build the sales page, payments, and install funnel to match. Work that used to take weeks now ships in a day.

Watch the Deal Engine race Chris's manual review

Land funding

AI implementation for your business

#seriouslandcapital #claudecode #landinvesting #aiautomation #productlaunch