AI
Big headlines in AI
There were lots of groundbreaking headlines this week, from new models and agents to chips. Something that particularly caught my eye was Project Mariner, an agent that works in your browser and is designed to complete tasks autonomously. It’s a promising tool that hints at the future of task delegation and productivity. I joined the waitlist for Project Mariner, which you can also do via Google Form on the site I linked above.
AI in the Wild: What Are Companies Using?
Over the past week, I’ve been looking for real use case examples of companies using AI. I’m particularly interested in companies that are not already in the AI business or operating in automation, etc. This week’s shoutout goes to legal teams using services that create templates and review contracts with AI. I believe the one I heard about directly was “DocVerify” however perhaps I misunderstood (or it was miscommunicated) as after I looked into it, it didn’t seem to actually be an AI tool. Perhaps someone I heard from just had the name wrong.
This Week’s Personal Project: Data Anonymization Tool
A major challenge that companies adopting AI face is addressing security and confidentiality concerns. It’s a huge hurdle that most companies are still grappling with. This inspired me to start building a Python-based tool to anonymize Google Drive files so that they can be used in training AI models. I got stuck early on with the automation aspect of this (it was over my head technically speaking), so I decided to chunk the project into pieces and focus on only accessing Google Drive files with Python, using Google Sheets as a test case. This step proved to be more forthcoming and rewarding as it was doable. I’ll be working on the anonymization piece in the coming weeks. An insightful part of this project involved getting to know Google Console which I'm excited to explore more in the future.
“Low-hanging fruit” is automation of Operations
Over the past two weeks, on my coding and AI journey, I’ve gained insights into what’s possible for one person to do with AI from a technical perspective to what financial limitations one can quickly run into using the LLMs. Starting small with a services use case feels like the right approach. For example, using natural language query to recall information from policies or putting a thin UX wrapper around a model that’s been trained on something specific. My next project will focus on “low-hanging fruit”— e.g., these simple, impactful automation solutions. Stay tuned for updates!
HR & PEOPLE TEAMS
People (HR) Topic: Performance Management
It’s the week before Christmas and at work we head into our year-end performance and compensation cycles after returning from the holidays (a process I largely influence). I’ve been thinking about what makes companies perform well and often go extremely deep down the rabbit hole of the root causes of performance culture or lack thereof. One common contribution to lack of performance is lack of motivated employees. And often management and FP&A contribute to that lack of motivation by not providing the right environment. I think back to Daniel Pink who laid out the ingredients for motivation - autonomy, purpose and mastery.
Lack of providing transparency, and ownership especially over budgets is a root cause to a sense of lack of autonomy by budget owners. If teams don’t have autonomy, they won’t be motivated and equally as painful their performance isn’t measurable because benchmarks / goals (e.g., $ amount of a budget) never get properly set and communicated. And at year end no one can really talk about how effective someone was with their resources.
Without clearly delegated financial and headcount budgets, organizations struggle to offer autonomy to their managers and P&L owners. Without this autonomy these leaders can’t take calculated risks. Companies need to let those animal spirits out, which will propel them to new heights.
CODING
Week 2 of Self-Taught Coding (Python) Practice
I’ve learned a lot this week, but the journey hasn’t been easy. The hardest part about starting to code is knowing where to go. Bootcamps often emphasize learning coding languages, but they don’t address essential questions like:
Where should I write my code?
What’s an API, and why is it bad to share?
How do I request permissions?
How does code translate into a user experience?
How do I create file paths that actually work?
Code Editors: A tale of two tools
Last week, I praised VS Code for its helpful UX. However, I’ve since discovered Cursor, a code editor that integrates AI to review your code directly within the interface. Cursor’s features streamline debugging and accelerate learning.
CLOSING INSIGHTS
The theme of employee enablement shines through for me this week.
Companies that don’t use AI will fall behind over time. More specifically, management that doesn’t enable the use of AI for everyone in their day to day work, will not be as competitive and dynamic as they can be.
Companies need their people to perform with or without AI and in order to do this they need to provide a sense of autonomy especially to budget owners.
I am envisioning and expanding on a new job emerging: Chief Automony Officer anyone? How cool would it be to have a job that focuses on ensuring people have all the right tools to be successful with AI and also ensure that management is providing the most autonomous and entrepreneurial environment possible?
The journey of learning and experimenting continues….