Welcome back to Jeremi and Luca’s Newsletter, an update every Sunday from two friends connected by a relentless desire to learn.
Jeremi: All In
I’ve started to become more confident about my skills and knowledge base in Machine Learning.
The more I learn, the more I feel like I can draw a cohesive picture in my brain. I have a language to describe what I’m seeing—both in terms of understanding the terminology, and having some intuition as to recurring themes and approaches to different problems. You might be surprised to think that “themes” exist in a technical field like this, but they do.
Part of this confidence comes from research papers that I read. A year ago, I joined a reading group and read a new paper each week. At the time, they were confusing and difficult to parse, and I was slow to get through them.
Since then, progressively reading more papers has helped improved this skill. Especially in the past few weeks, I’ve been reading a lot of research papers as I figure out which direction I want to take my class project and research project.
I’m at the point where I feel like I can pick up a research paper, read through it, and understand most of the content (at least, in a specific area of Machine Learning where I have more experience).
Two other anecdotes:
Last Tuesday, I went into my professor’s office to propose an experiment for our research, based on papers I had read. After some discussion, he agreed that it was a solid plan and I’m moving forward with it.
On Saturday, I was up in the Bay Area. I went to a networking event in SF, where I spoke with an AI safety engineer about the research I was working on, the research she was doing, and different developments that were happening in the field.
These are all small examples. But they feel meaningful. Being able to read and understand papers as they’re released. Having peer conversations with people who are active in the field. Coming up with my own ideas and pursuing them. I’m transitioning from a spectator to a player. And that feels exciting.
No doubt, I have much, much, to learn. I know just enough to sound smart, and not enough to realize how much I don’t know. If I were to put myself on the graph of the Dunning-Kruger effect (depicted below) I might be somewhere on the “Peak of ‘Mount Stupid…’”

Still, it’s nice to see the compounding effect of spending a lot of time working in AI. It makes me want to dig my feet in and go even deeper.


Luca: Diversification
This week’s entry will just be a few bits and pieces from the week…
Saturday, flew down to Albuquerque to meet my mom who was driving her late dad’s belongings across the country from North Carolina. A childhood best friend joined her up until New Mexico (where she lives), then I took the last shift.
Sunday started in New Mexico and ended in California, but most of the day was spent driving through Arizona. We stopped in Flagstaff for a pizza place called Fat Olives that’s ranked in Yelp’s “Top 100.” We can confirm that the pizza was very, very good.
On Monday, we drove from Barstow, CA back up to the Bay Area. On the drive, I had a call to tour a summer apartment in New York City—and locked that in a few days later. I’ll be living with another intern right across the Manhattan Bridge in Brooklyn.
Tuesday was pretty school-focused, but I broke for lunch with a senior at the business school who’ll be working at Shield AI after he graduates. Shield is a super cool defense tech company, so it was interesting to hear about that.
The rest of the week, I focused heavily on a technical project I needed to submit for a “Deployment Strategist” internship at Palantir. The program would be in New York and would replace my fall semester at Cal. Palantir’s whole goal is to have people leave school to keep working after their internship ends.
Already, I was hesitant. The dates didn’t work out well and I certainly wasn’t ready to commit to not returning to school in the fall.
But the company and role absolutely interested me. On top of appreciating their mission and reconstruction of the consultant role, I spoke to a law professor here at Berkeley who got me excited about the types of analysis possible with the company’s tech: layers deep supply chain analysis for aircraft parts, data-driven understandings of criminal networks and their various nodes, etc.
And anyway, I figured completing the technical project would be valuable to me either way. Thus, I’ve been working on the application for the past few weeks.
The deadline rolled around last Friday, and I wasn’t finished with the project. By then, I had pretty much decided that I didn’t want to do the internship, and I think that feeling contributed to me not putting as much effort into the project as I could have.
But I wish I decided this earlier and more intentionally. I got myself to a point where there was no other option; I closed the door on myself. No matter what the right decision was, I didn’t make it of my own accord and that disappoints me.
I was right, though, about one thing: the technical project was valuable either way. I learned a lot, so no time was wasted. And the work I’ve done will always be there if I ever decide to apply to another role at the company.
I wrapped up the week by finally finding someone to sublease my Berkeley apartment this summer, attending sorority “date party” with Caroline, eating dinner at home with my family, doing some grocery shopping, and getting ahead on next week.
Luca: I can’t thank you enough for taking your time to accompany me the rest of the way back home last week. I will treasure that time with you always! Best son ever!! 🥇🏆
Jeremi and Luca, you continue to amaze me with your thirst for knowledge and your growing ability to hone your skills and challenge yourself. Your intellectual world is so much greater than mine was when I was in college - your pursuit of excellence gives me a greater measure of hope for our future, when your generation comes into its own. Thank you for sharing.