- Operators
- Posts
- 022 AI for Operators
022 AI for Operators
Szilvia Kocsy, former Chief of Staff at LinkedIn; 10 links
Tom Guthrie
November 05, 2025
Hi there,
Welcome back to AI for Operators. Here’s what we’ve got for you this week:
The Operator: Szilvia Kocsy (former LinkedIn and PWC - listen here)
The Essay: The AI Dividend
The Links: 10 links, including the first AI-native NBA team, AI systems for Chiefs of Staff, and how AI is changing the corporate real estate industry.
The Operator |  |
This week’s episode is with Szilvia Kocsy, Founder and Global Chief of Staff & PMO Consultant at Life Mentor Global, former Chief of Staff to the Managing Director of EMEA & LATAM at LinkedIn  |  |
Szilvia Kocsy, Consultant and former Chief of Staff at LinkedIn, shares how operators can navigate AI’s pace, add guardrails, and prepare for career shifts. We discussed:
AI best practices are still early stage: ethics, data controls, and usage policies remain immature; leaders should pilot, document guardrails, and iterate as capabilities evolve.
AI democratizes production, but audiences still seek human perspective and connection; combine human spark with AI scale for resonant outputs.
70% skills shift by 2030: massive change is inevitable; prioritize continuous learning.
Shadow-AI risk management: choose approved tools, set confidentiality rules, and create team champions and hackathons to spread safe, effective habits.
Helping executives gain leverage: use AI to stress-test decisions, mine meeting transcripts for patterns, and tighten enterprise messaging. The result? More clarity in less time.
Workflow rebuilds beat small changes: map current processes, then automate, augment, and amplify.
Put your “mask” on first: automate emails/records/meetings, use AI to strengthen judgments, then focus human time on strategy. Once you’ve tested the processes, then cascade them with your CEO and team.
The power of community: build peer learning (e.g., Szilvia’s Life Mentor Circle) to accelerate upskilling and normalize shared AI frameworks across functions.
The Essay: The AI Dividend |  |
In international relations, the concept of a “Peace Dividend” is used to explain the second-order economic benefits that come to former combatant countries after they lay down their weapons. There’s a similar second-order benefit waiting for companies that start preparing in earnest for AI.
Certainly, the benefits of implementing AI systems can be large: higher production, more efficiency, lower costs. But fully implementing AI can seem like a daunting task, and those benefits can therefore seem remote.
What’s underappreciated is that simply preparing to implement AI will give your organization lasting, sizable rewards that may outweigh the direct perks you’ll see from implementing AI.
How?
Successful AI implementations will go through a four-step process before they launch a new solution into their team:
Data Mapping: Building a comprehensive view of all the data flows and depositories in your organization.
Process Identification: Creating a compendium of all of the process that go on each day/week/month in your organization.
Interviews: Get in the weeds with folks on the ground to identify bottlenecks, opportunities, and perspectives.
Process Mapping: A full playbook of all of the processes identified in Step 2.
Done right, this is a lot of work. But the benefits are real.
Identification of waste and inefficiencies: once you know all your processes and data flows, you’re guaranteed to find a few that can be cut.
A backlog of ideas: I dare you to dive this deep into the inner workings of your org and not come out with a laundry list of things you want to work on.
Buy-in from your team: because what’s more persuasive, someone who bothered to ask you about your perspective and has a seemingly comprehensive view of everything going on at the company, or someone who didn’t bother and doesn’t have that view?
A more resilient organization: someone quits or goes out on leave? That’s okay, you know all the processes that they’re involved with and can onboard someone new with your step-by-step playbooks easily.
So, even if you (or your exec team) are skeptical about whether it’s the right time to go all in on AI, do the preparation work anyway. You’ll reap the AI Dividend.
The Links |  |
Practical
When should you use an agent vs. a subagent vs. a tool? Aaron Levie shared how he thinks about it.
More Aaron Levie on how AI is changing product prototyping for PMs.
How to get started with Claude Code - the most practical answer
AI Systems for Chiefs of Staff: a great presentation shared by a future AI for Operators podcast guest at Chief of Staff Network’s annual NYC summit
Perspectives
Peeling the AI Anxiety Onion: anxiety about AI can be the thing that stands in the way of your organization fully adopting it. The biggest predictor of anxiety? Whether the worker uses AI already or not (those who don’t are more anxious - rightfully so!).
One developer’s perspective on how AI is changing the corporate real estate industry. What’s the version of this for your role/industry?
The San Antonio Spurs - the first AI-native NBA team?
News
Applied Compute raises $80m to build enterprise AI agents for specific companies: specific intelligence > general intelligence
New jobs created by AI - WaPo has 16 examples
Thanks for reading,
Tom Guthrie