[Frontier] Update: Summer '24
lots have happppened
It has been a while since I’ve written here. The semester is nearly finished, and a lot has changed. So, I thought it was time for a quick update.
I’ve launched three projects, learned a lot about developing applications, and decided to take next semester off.
I did not find much joy in following school curriculum1. I wanted to learn about things that truly interested me. Had enough of solving textbook problems and memorizing things I would forget in a week. Plus, had a hard time finding inspiring colleagues on campus2.
I’ve learned a number of things about myself on this journey. But I still crave the 100-hour work week. I’ve never put everything I had into a single goal since I entered uni: always doing side gigs, chasing multiple goals.
It’s time for me to double down on something. I’ve seen friends meticulously working on their projects for 2–3 years. They are the ones showing real progress: building companies that are creating revenue, and publishing frontier research papers.
I wish I could tell you everything that was going on behind the scenes, but I think that’s for another time. I’ll leave you all with a collage of my recent thoughts:
The days are long, but the decades are short: half of ‘24 is gone. It’s been 4 months since I was discharged, 4 years since I entered uni, and 8 years since I lived abroad (absolutely crazy).
User-generated (multimodal) agents will fuel the next big thing in B2C. The thing no one has figured out is the form factor.
Software that helps engineering teams figure out how to fine-tune (with or without training) large models at a low cost will become huge.
Stay close to people doing their projects, and you should be working on yours as well. I see an alarming lack of hackers and an abundance of conformists on campus :( Do your thing, and build the future!
Don’t spend too much time networking. You won’t learn as much as you do by building. And you’ll eventually meet great friends on the way if you build something of value.
If you’re not technical, learn how to code. You can do it in a month, or even in 2 weeks. It’ll give you superpowers (take my word for it).
Engineering is the skill needed to build venture-scale startups, especially as undergrads. We don’t know sh*t about anything (with a few exceptions). So the only solution is to study, build, launch, and iterate: engineering is required in the process.
Choose your quests wisely. If you’re at level 1 and go after a level 100 quest or a level 3 quest that takes 5 years to finish, it’s very likely you’ll quit along the way.
Work with responsible people.
Work with people who have aligning long-term goals.
Work with people who give you surprising insights.Best video I’ve watched this year: “How to Live an Asymmetric Life” by Graham Weaver. Never say: ‘Not me, not now.’ The antidote to fear is
Do hard things
Do your thing
Do it for a decade
Write your own story
All that matters in life is finding meaningful work and finding love (romantically / with friends and family). Then, let them kill you. (you’ll die without them anyway.)
The problem is, how are you going to find both? I personally haven’t figured it out yet. Again, I feel like the only solution is to look back at yourself, then launch & iterate.
Thanks for tuning in :) Hope you guys have a great summer.
In fact, I’m writing this while procrastinating on studying. If I do end up returning to school and taking my studies seriously, it will be to pursue frontier research.
Ambitious people doing interesting things did not put much effort into school. They were either working on startups or in the lab crunching experiments. Please do reach out if you are a hacker at heart.