- Operators
- Posts
- 024 AI for Operators
024 AI for Operators
Jill Czarnik, Head of Enablement at Wordly, You're Not Making Enough Money, 10 links
Tom Guthrie
November 19, 2025
Hi there,
Welcome back to AI for Operators. Here’s what we’ve got for you this week:
The Operator: Jill Czarnik (Head of Enablement at Wordly - listen here)
The Essay: You’re Not Making Enough Money
The Links: 10 links, including a Claude-based cyberattack (and how to replicate it), a profile of Greg Brockman, and the problem with pilots (no, not the ones at JFK).
The Operator |  |
For this week’s episode, we spoke to Jill Czarnik, Head of Enablement at Wordly and former Chief of Staff & Head of BD and Strategic Growth at Persado.  |  |
Jill Czarnik, Head of Enablement at Wordly AI and multi-time Chief of Staff, shares how she drives AI adoption with a $0 budget.
Work with what you’ve got
Exploit vendor roadmaps: stay close to launches, test early, and collaborate with their product teams (examples: HubSpot prospecting agent maturing; Vanta AI add-on streamlining security questionnaires)
Audit what you have before buying new: HubSpot’s AI Workspaces proved to be useful for prioritizing pipelines and automating follow-ups with no new spend (but some time investment from current sales reps)
You can build a scrappy outbound engine: Octoparse for prospecting, Clay for enrichment, Gemini to personalize templates, Hubspot for execution - much cheaper than ZoomInfo and other tools
The reality check: expensive AI platforms often sit idle (or at least are underutilized) - be the operator who either drives adoption or cuts waste and captures savings
Focus on adoption
If leadership model usage and you create ways for ICs to showcase their wins, you’ll get more organic adoption of the tools
Focus on durable workflows: avoid party-trick automations; apply the 80/20 lens to prospecting, pipeline hygiene, and communications that compound weekly.
Scaling yourself
With a full-time job and a side hustle as a fractional CoS, time is Jill’s scarce resource: she sends AI note-takers to some meetings, reviews summaries, and acts on decisions instead of joining meetings live
The Essay: You’re Not Making Enough Money |  |
Over at Chief of Staff Network, we just ran our annual compensation survey (check it out here if you haven’t already). And while the data from the hundreds of respondents had many interesting takeaways (women earn more than men! average salary reached $167k!), there was one statistic that really stopped me in my tracks.
It’s that AI maturity - both for the individual and the organization - have a huge impact on how much Chiefs of Staff are getting paid.
For example, AI-First Chiefs of Staff are getting paid 28% more than AI-Novice Chiefs of Staff. Similarly, Chiefs of Staff at AI-First organizations are getting paid 23% more than Chiefs of Staff at AI-Novice organizations.
The fact that both of these premiums exist suggests that companies recognize the value of Chiefs of Staff with AI skills (that’s the whole premise of this newsletter, so we agree!), and that companies with AI capabilities are doing well enough to pay higher prices to talent to keep that advantage going.
However, if you want to command the highest salary, don’t go to a company that’s AI-First. Instead, be an AI-First Chief of Staff at an AI-Curious organization (the lowest level in our survey). If you’re in that situation, you’ll get a salary that’s a whopping 47% higher than the average Chief of Staff salary.
I think of this as the “transformation premium” - the CEO or executive team recognizes that they need to evolve into an AI-First organization, but they don’t have the capabilities in house. They can either hire Accenture or another consulting firm, or they can pay a Chief of Staff to lead the charge (and some are likely doing both). I know which I would bet on.
So, if you’re looking to give your career a boost, don’t just learn the AI skills, go put them to work. And if you’re at an organization that knows it needs to accelerate its AI capabilities, get in touch with us - we have a network full of AI-First Chiefs of Staff that we’re training to do just the kind of transformation that your organization needs. We’d be glad to figure out a way to help out.
The Links |  |
Practical
Are Pilot Programs the Problem? This HBR article argues that focusing your AI firepower on single problem - and insisting on getting it right - is better than trying to go an inch deep on everything.
Give Your AI a Job Interview: Ethan Mollick makes the case for rigorous real-world evaluations of AI models, if you’re going to rely on them for real-world tasks like you would a colleague. Makes sense to me.
AI Coding for Complete Beginners: If you’ve been putting off learning how to use AI to code, here’s a good place video tutorial to start with.
Looking for inspiration? This list of Claude use cases from Anthropic is a good place to start.
Don’t Stop at Second Brain - Build a “Critique Crew”: A PMM at Descript shows how he helped a colleague automate the collection of more diverse perspectives in order to improve design reviews. Relevant for many tasks, from board meetings to marketing campaigns!
Perspectives
Some useful charts from a16z: Insights include that agent power-users (6% of the survey respondents) saw a 5% EBIT impact from their usage - that’s real ROI (if we assume it’s positive)!
Meta Adds “AI Impact” to Performance Reviews: Nothing promotes adoption like incentives (aka putting someone’s job on the line). While this won’t help in allaying anyone’s fears about impending job losses from AI, these kinds of actions are what creates cultural change at companies.
News
The First (Reported) AI-Orchestrated Cyberattack: This overview from Anthropic is relevant for a few reasons: 1) Know that these types of attacks are coming and prepare, 2) Anthropic lays out how the hackers architected the system to execute the attack (you can use a similar architecture for legitimate ends!), 3) Anthropic reports that one of the reasons that the attack was not more successful was that Claude Code hallucinated so-called ‘secret’ information, including fake employee logins. Even in a sophisticated, highly-engineered system, these LLMs will still produce fake outputs, so you must have a tightly-controlled process to verify the veracity of what they produce!
A profile of Greg Brockman, the non-Altman “power broker” behind OpenAI’s maximalist ambitions. Notable details: reports on his demanding management style, his work on everything from technical details to lightning-speed multibillion-dollar deals, and his political donations.
Scribe Launches Optimize: previously covered in AI for Operators, Scribe is an incredibly useful tool for folks to document workflows in order to get AI ready. Optimize tells you which workflows to start with, how to improve (including ROI), and a step-by-step plan to get there. Can you say business transformation in a box?
Thanks for reading,
Tom Guthrie