POWERS Podcast #392 Summary

Guest: Jason Baxter, Co-Founder @ Fostr AI
Host: Chris Powers
Date: September 4, 2025
Theme: What Elite CEOs Know About AI That Others Don’t

Episode Overview

This episode dives into how elite CEOs and executive teams are realistically implementing and leveraging AI in their organizations—shifting the focus from shiny chatbots and one-off agents to building foundational, business-specific infrastructure like knowledge graphs. Chris and Jason unpack why 99% of business AI is about data context, not just adopting popular tools, and how Fostr is helping companies create and own their proprietary knowledge engines for a compounding strategic advantage.

Key Discussion Points & Insights

1. How Executive AI Thinking Has Evolved

• CEOs are overwhelmed by AI’s rapid evolution and firehose of information.

  • Most try to solve a single problem, missing the broader opportunity.
  • “Every person that's using AI has a view on how they're going to use it based on what little information we've all learned about AI right now.” – Jason (04:32)

• Mainstream tools (like ChatGPT) shape shallow thinking.

  • Executives incorrectly equate AI with generic chat interfaces.
  • “CEOs really need to...understand what is actually possible in these companies, not through the lens of these large language models.” – Jason (07:15)

2. The Real AI Opportunity: Owning & Compounding Knowledge

• AI as a utility: Models will be abundant and commodified, but data context will be king.

  • “[LLM] models now are becoming a utility and they're improving every day ... but that's not the business that's going to solve your company's issue.” – Jason (08:14)

• Infrastructure, not interfaces, is the moat:

  • The companies winning with AI are not reselling chatbots, but building robust knowledge engines tied to their business.
  • “What has to happen for people to truly leverage AI is the infrastructure has to be built. And that infrastructure is where we play.” (08:48)

3. Universal AI Implementation Problem: Disorganized Company Data

• Almost every company faces the same root issue:

  • Messy, unstructured, siloed data makes AI almost useless at scale.
  • “No company’s trying to go straight to autonomous AI agents… What they're saying is ‘how do I, I have this data over here, and I have this information over here. I would love to…combine that data.’” – Jason (10:45)

• Band-aid approaches with chatbots only amplify chaos.

  • “Every day they do that...without the compounding of the knowledge…they're actually making the problem worse.” – Jason (13:01)
  • Firing isolated AI tasks with messy data just “escalates chaos.”

4. First Principle: Structuring Company Data for Context

• The elegant solution: a continuously updating Knowledge Graph.

  • Foster’s core innovation is automating the ingestion, structuring, and relational mapping of all company knowledge.
  • “What you get when you do that is a graph…nodes of information that are all connected to each other. That is critical when you want to interface with AI…” – Jason (14:27)

• This knowledge graph becomes the business’s real memory and brain.

  • “It is the compounding knowledge of an entire organization, living, breathing organization, that is growing all the time. That is what AI is in the future.” – Jason (00:00 & 30:32)

• Specific examples:

  • Integrating Slack, Zoom, Notion, Airtable, Google Drive, Email, Calendar, etc. into one evolving knowledge graph.
  • Companies move from siloed data to dynamic, context-rich organizational memory.

5. How the Knowledge Graph Compounds Value

• Value comes not just from aggregation, but from Q&A cycles.

  • Every interaction (a question and answer) feeds back into the knowledge graph, so it gets 10x smarter over time.
  • “It isn’t the value of a knowledge graph in the data coming in, it’s in the questions that get asked ... and the responses.” – Jason (36:30)

• Owned knowledge graphs will be the new strategic assets ("the new gold").

  • “That knowledge graph becomes a proprietary model of the company that no one has, except that company.” (37:48)
  • This enables eventual creation (or licensing) of specialized, private LLMs tuned to one’s business.

6. Why General-Purpose AI is Not Enough

• Public LLMs (OpenAI, Grok, etc.) will never understand your business as well as your own knowledge engine.
  • “More knowledge of the entire world is not going to help your business, right? What your business needs is specific knowledge.” – Jason (41:30)
  • Relying on general models means you’re training them, not building your own advantage.
  • “Who got the benefit of that? You think you did … but who really got smart was ChatGPT. Because ChatGPT owns that question and it owns that response.” – Jason (46:40)

7. Data Ownership & Security

• Fostr as a multi-tenant, private cloud:

  • “We did that specifically so that every single user has its own instance … has a private server.” – Jason (49:45)
  • Foster provides the rails, but companies retain ownership and control of their knowledge graph.

• Future market: knowledge graph provenance and validation

  • “If you have a proven model, proven model means there is an actual company on the other side of it with proven success … those are the models that will be more valuable.” – Jason (52:49)

8. The Coming AI Disruption (and Survivor Profile)

• Companies who resist won’t survive.

  • “If you don't start down this path, your business will not look like it does today in 10 years...this transition is happening.” – Jason (54:47)
  • AI is now automating even highly-complex tasks like building data pipelines—what took engineering teams “months…can now be done in…minutes.” (34:24, 56:33)

• Survivor profile:

  • Tech-savvy, digitally native, with a strong AI-first leadership mentality.
  • Early adopters with clean(ish) data will have a massive compounding advantage.
  • “Young digitally native companies…will skyrocket past most companies because they won't know any different.” – Jason (24:07)

Notable Quotes & Memorable Moments

• “It is the compounding knowledge of an entire organization, living, breathing…that is growing all the time. That is what AI is in the future.” – Jason (00:00, 30:32)
• “Every person that's using AI has a view on how they're going to use it based on what little information we've all learned...the noise is so strong right now.” – Jason (04:32)
• “What has to happen for people to truly leverage AI is the infrastructure has to be built...not the flashy new toy.” – Jason (08:48)
• “You would have to literally go to every individual and say, please start sharing with me all your files.” – Jason (27:02, on messy data)
• “If you lump yourself into that [public LLM], you are not creating advantage. You are getting sucked into a hole…they are going to own your business or your business model.” – Jason (48:13)
• “Knowledge graphs themselves are going to become the new gold of the world…proprietary to the company that compounds it.” – Jason (37:51)
• “Used to, you would have to pay a data engineer or a team…to map data from one system to another…AI has solved the ability to map that data and ingest it like that.” – Jason (56:34)

Important Timestamps

• [00:00] – Jason on what AI really is (compounding, living company knowledge)
• [04:32] – CEOs’ struggles and the strong noise in the adoption journey
• [10:45] – The universal problem: disorganized company data, and why most “agent” talk is (currently) marketing hype
• [14:27-19:20] – Why structuring data and building a knowledge graph is foundational
• [22:27–26:27] – Realistic requirements for companies to get started (messy vs. clean data & why “AI-first mentality” is more important)
• [30:32-34:00] – Why a knowledge graph is the only way for AI to truly add value in a business
• [36:30] – The real compounding value: not just the company data, but the questions and answers generated over time
• [46:40] – Who really gets smarter when you use a general AI model (spoiler: not you)
• [49:45] – Foster’s data model: privacy, control, and customer data ownership
• [56:34] – The new AI capability: complex data engineering tasks now automated in weeks
• [57:50] – Fostr’s upcoming public launch & onboarding strategy

Episode Structure (Section Guide)

1. What CEOs Are Missing About AI (00:00–09:56)
2. AI in Practice: Infrastructure, Not Hype (10:32–18:00)
3. The Data Problem and Knowledge Graph Solution (18:00–34:00)
4. Knowledge Graphs: Proprietary Memory & Advantage (34:00–44:00)
5. Q&A: Licensing/Monetizing Knowledge, Data Ownership, Future-Proofing (44:00–56:00)
6. How Fostr is Rolling Out: Public Onboarding & Next Steps (57:30–end)

Takeaways for Listeners

• Don’t think “implement AI,” think “build an internal knowledge graph.”
• Focus on infrastructure—compounding and organizing your knowledge—not just tool-tinkering.
• The strategic edge in AI will come from owning, growing, and leveraging your company’s unique knowledge engine—not from relying on the same public models as everyone else.
• The future will belong to organizations that start piping in their data, structure it for context, and ask increasingly valuable questions—allowing proprietary models and insight to emerge, setting them apart in their industry.

For More:

• Early access to Fostr: FostrAI.com
• AI CEO Network (Slack community): ceonetworkosterai.com