Silicon Valley GirlChief AI Architect: How to Make AI Your Strategic Partner in 40 Minutes | Conor Grennan
CHAPTERS
- 0:00 – 2:00
Why most “AI adoption” stats are misleading: power users vs casual users
Conor explains that many people say they “use AI,” but their usage is shallow—similar to claiming Excel proficiency without knowing formulas or doing real work in it. He defines a true power user as someone who uses AI across home and work, frequently, and in long conversations.
- •Adoption looks high, but meaningful adoption is much lower
- •Analogy: Excel users vs Excel power users
- •Power use = cross-context use (home/work) + high frequency
- •Long conversations beat one-off queries
- •AI should be treated as a companion, not a search box
- 2:00 – 3:35
A simple way to start today: apply one model to the tasks you already do
Instead of chasing new tools and prompt libraries, Conor advises picking one AI tool and using it to augment daily tasks. The goal is to reduce FOMO and build a consistent habit of asking for help throughout the day.
- •Don’t stress about being behind or constant tool releases
- •Avoid prompt-library FOMO; focus on everyday tasks
- •Pick one tool (ChatGPT/Gemini/Claude/Copilot) and start
- •List your recurring tasks and ask how AI can assist
- •Think of AI as “help every day,” not a PhD-level system
- 3:35 – 4:59
Game-changing everyday use cases: scheduling, writing, and idea-generation
Marina shares a practical automation win—turning emails into calendar events—while Conor highlights creative and strategic use cases. His biggest “life-changing” value is brainstorming and iterative improvement for writing and teaching.
- •Email-to-calendar automation as a high-ROI daily workflow
- •Conor’s top use: brainstorming via back-and-forth dialogue
- •Write your first draft yourself, then use AI to improve it
- •Ask for readability improvements and audience-fit edits
- •Use AI to surface stats, gaps, and stronger framing
- 4:59 – 6:27
The key mindset shift: AI as a collaborator that replaces “a million things”
Conor argues organizations mistakenly treat AI like a typical digital transformation (old tool swapped for new tool). Generative AI isn’t a single replacement; it’s a new way of thinking that can support nearly any task, from work strategy to personal planning.
- •Digital transformation ≠ AI transformation
- •AI doesn’t neatly replace one tool (not just Google)
- •Better mental model: a new person sitting next to you
- •Use cases are broad; over-focusing on narrow ones can limit value
- •AI can help with thinking, planning, and problem-solving across life
- 6:27 – 9:14
Why your brain treats AI like Google—and how to rewire that habit
Because chat interfaces resemble search engines, people default to “command → response → leave.” Conor explains this is a brain-level automation issue and that users need deliberate behavioral frameworks to create conversational, iterative interactions.
- •Interface similarity triggers “search mode” in the brain
- •Search behavior: command-response-walk away
- •Conversation behavior: iterative dialogue like texting a friend/coach
- •Habit change parallels fitness: knowing what to do isn’t enough
- •Frameworks help overcome default neural pathways
- 9:14 – 10:57
Stop chasing tools: focus on process redesign (AI is a “process machine”)
Conor advises that most people don’t need to track every model update. The bigger unlock is mapping your workflow and inserting AI into each step, like a CEO evaluating a new office location or a team building a slide deck.
- •Most people don’t need the newest model versions
- •Tool choice matters less than process improvement
- •AI is not an “answer machine”; it accelerates workflows
- •Break work into steps; ask AI to support each step
- •For slides: use AI for structure/visual ideas, then build yourself
- 10:57 – 13:31
Which tools he uses—and why general LLMs often beat niche apps
Conor shares his rotating set of core models (Claude, Gemini, ChatGPT, Copilot). He argues that comfort and consistency matter more than small performance differences, and that building a “companion” habit is more valuable than chasing specialized tools.
- •Claude (writing strength), Gemini (deep thinking + Google integration)
- •ChatGPT as everyday quick-reference; Copilot for enterprise contexts
- •Performance debates can be overblown for most users
- •Choose what feels natural for your workflow
- •A consistent “companion” behavior outlasts tool fads
- 13:31 – 16:44
Portable “AI memory”: how to transfer your context between tools
To avoid starting over when switching models, Conor describes creating a distilled personal/business dossier from ChatGPT’s memory. You can paste that document into other tools so they quickly understand your preferences, goals, and working style.
- •Use one model to summarize everything it knows about you
- •Ask for a distilled multi-page ‘profile’ (preferences, style, goals)
- •Paste that into Claude/Gemini to bootstrap context
- •Create multiple “co-CEO” instances across tools
- •Memory + personalization dramatically improves relevance of outputs
- 16:44 – 18:10
Power prompts that upgrade thinking: “Push back” and “What am I missing?”
Conor’s most valuable prompts are those that force critique and iteration. He recommends explicitly instructing the model to challenge you and repeatedly ask for gaps, improvements, and second-order considerations.
- •Tell the model: “Push back” to reduce complacent outputs
- •Use “What am I missing?” repeatedly to deepen quality
- •Iterate beyond the first answer; don’t stop at draft one
- •AI can help you overcome mental fatigue around revisions
- •Editing your own work with AI often beats AI-first drafts
- 18:10 – 24:10
Accuracy, hallucinations, and when not to trust AI
Conor reframes hallucinations as a known limitation similar to human error, emphasizing that AI isn’t a calculator. For high-stakes topics (medical, financial, safety), he recommends verification via primary sources, manuals, or trusted experts.
- •AI is not deterministic like a calculator; treat it like a smart assistant
- •Humans are wrong too—so adopt a verification mindset
- •Use AI for brainstorming; verify for high-stakes precision
- •For critical decisions: go to source documents and expert input
- •Don’t rely on AI for safety-critical instructions or exact facts
- 24:10 – 26:34
Critical thinking and education: AI can weaken or strengthen learning
Conor and Marina discuss the risk of outsourcing thinking to AI. He compares it to CliffsNotes: it can either replace real learning or deepen understanding through explanation, alternate perspectives, and personalized help.
- •AI can erode critical thinking if used only for answers
- •Used well, it becomes a personalized learning assistant
- •Parallel to CliffsNotes: shortcut vs deeper comprehension tool
- •Education must adapt to guide productive AI usage
- •The main variable is behavior, not the technology
- 26:34 – 28:29
Skills that matter when nobody knows the future: behavior, fluency, and reinvention
Conor rejects vague advice like “be curious” and focuses on actionable behavior changes. The differentiator will be people who use AI to reinvent workflows in their domain and spread that innovation across teams and organizations.
- •The future is uncertain; leaders don’t fully know either
- •Curiosity can be treated as a behavior with steps and routines
- •Domain expertise still matters because it defines ‘quality’
- •Bottom-up reinvention will drive real organizational change
- •Upskilling should target whole teams, not isolated individuals
- 28:29 – 30:44
Hiring disruption and the entry-level squeeze—plus how job seekers can stand out
Conor outlines why entry-level white-collar work is most exposed and cites the possibility of large reductions even with today’s tech. His job-search advice: demonstrate how you’d redesign the role and the surrounding team workflows with AI, not just claim you “use AI.”
- •Entry-level tasks are most vulnerable to automation leverage
- •Even without major tech advances, productivity gains can shrink teams
- •Big problem: fewer entry-level tasks means fewer training grounds
- •Standing out: show a concrete AI workflow for the role
- •Best signal: redesign the team’s process, not just your personal output
- 30:44 – 35:23
Why it may be the best moment to be an entrepreneur (including inside a company)
Because AI can act like a “synthetic team,” individuals can build and test ideas faster than ever. Conor emphasizes that reinventing processes gets noticed more than simply doing more work—and can become repeatable IP you can scale inside a company or into a business.
- •AI lowers barriers: coding/marketing/ops support via ‘synthetic teams’
- •Build a ‘parallel resume’: job + side venture experiments
- •Opportunity lies in repeatable, expandable process solutions
- •Reinventing processes creates visible ROI and career leverage
- •Entrepreneurship can be internal: becoming a process innovator leader
- 35:23 – 37:29
Universities won’t disappear—but online learning and the middle tier will change
Conor argues universities are more than content delivery: they provide community, identity formation, networks, and experiential learning. However, AI will likely replace much of transactional online coursework and pressure mid-tier institutions.
- •Universities persist due to social and developmental value
- •AI can replace purely transactional content delivery
- •Online courses are especially vulnerable to AI tutoring
- •Likely consolidation: fewer institutions in the ‘middle’
- •Education shifts toward personalization and mentorship models
- 37:29 – 40:37
AI’s most meaningful impact: education and healthcare access in Nepal
Conor shares his work teaching AI in Nepal and why this is the change he most wants to see. He describes AI as a scalable tutor and a partial medical proxy in remote areas, expanding access where teachers, tutors, and doctors are scarce.
- •AI as an always-available tutor for underserved students
- •Potential to accelerate national progress via education access
- •Healthcare triage support for remote communities with mobile phones
- •Acknowledges limitations and need for verification
- •Frames AI as a humanitarian lever beyond corporate productivity
