I Lived Through Y2K. AI Is Bigger. | Ex-Microsoft Leader Carmen Aviles Builds Agents for Businesses
カートのアイテムが多すぎます
カートに追加できませんでした。
ウィッシュリストに追加できませんでした。
ほしい物リストの削除に失敗しました。
ポッドキャストのフォローに失敗しました
ポッドキャストのフォロー解除に失敗しました
-
ナレーター:
-
著者:
She lived through Y2K, sold for Microsoft, and bet early on the cloud. Now Carmen Aviles is building AI agents on a Mac mini, and she says this wave is bigger than all of them.
Carmen Aviles has spent more than 40 years in tech. Her path runs from shipping e-learning on CDs across Latin America to SAP implementations during the Y2K panic, Microsoft's partner ecosystem, and one of South Florida's first Office 365 practices. Today she's the founder of Aviles AI and helps small and midsize businesses adopt AI without breaking what already works.
Hosts Samantha Viola Wilson and Brittany George ask her what Y2K can teach us about the AI rush and whether AI will really take jobs in the next five years. They also get into how CEOs can get their teams to adopt it: role by role, with incentives, starting from the top. Carmen gives an honest account of building a six-agent real estate marketing system, including the cron job that burned $135 in one day. She explains how Jake Van Clief's Interpretable Context Methodology (ICM) brought her cost down to about $10 per run.
IN THIS EPISODE:
- How is the AI rush like Y2K, and how is it different?
- Will AI replace jobs in the next 5 years?
- How do you get employees to adopt AI? (role-based rollout)
- Is Microsoft behind in AI? (Copilot, Copilot Studio, GitHub)
- How do you cut AI agent token costs?
- Cloud vs. local AI models: who owns your data?
- Using NotebookLM as a personal AI tutor
CHAPTERS
00:00 Meet Carmen Aviles ("Abuela")
00:53 Shipping e-learning on CDs across Latin America
02:23 Y2K, SAP, and the panic that looks like today's AI rush
05:16 Getting hired at Microsoft (via cassette voicemail)
07:11 Betting early on Office 365 and the cloud
09:40 First AI "wow": 2-hour research cut to 30 minutes
10:37 The OpenClaw obsession and a 6-agent marketing system
12:13 Y2K vs. AI: what's the same, what's different
13:40 Early adopters, proof-of-concept people, and skeptics
16:26 Will AI take jobs? Carmen's 5-year outlook
18:32 Kids, critical thinking, and AI in education
20:35 NotebookLM as a personal AI tutor
23:22 How CEOs get teams to adopt AI
25:28 Free Microsoft AI courses and the partner opportunity
27:55 AI adoption has to start at the top
29:30 Is Microsoft behind in AI? Copilot, models, GitHub
34:32 Copilot Studio inside Teams: worth $30 a month?
37:16 Token costs and the $135 cron job mistake
40:26 Building vs. monetizing: the honest update
43:09 ICM: the folders-and-files method that cut costs
46:53 Protecting your IP when vibe coding
48:07 Obsidian, Skool, and knowledge graphs
50:47 Cloud vs. local models: who owns your data?
54:44 AI glasses and DIY robots
58:13 Personal AI: ChatGPT for life, Claude for business
1:00:44 How to reach Carmen
RESOURCES MENTIONED
- OpenClaw
- Interpretable Context Methodology by Jake Van Clief: https://arxiv.org/abs/2603.16021
- Clief Notes community on Skool: https://www.skool.com/cliefnotes
- Google NotebookLM
- Microsoft Copilot Studio
- GitHub
- Obsidian
- Brilliant Labs Halo AI glasses
CONNECT WITH CARMEN AVILES
Business Optimization & Adoption Consultant | Founder, Aviles AI
🌐 https://avilesai.com
💼 LinkedIn: https://www.linkedin.com/in/carmenavilesaiconsulting/
📸 Instagram: @caviles812
✉️ carmen@avilesai.com
🏡 Buying or selling residential or commercial real estate in South Florida? Carmen is a licensed realtor. Reach out by email.