Notes from the Zurich AI Conference 2025

These are notes from when I went to The Zurich AI conference 2025. This was the first time I've attended a conference solely for networking. Here's what I learned/experienced:

The people

  • Of the ±200 people that attended, I interacted with very few product/startup people, researchers / engineers, and especially no engineers from big tech.
  • There were primarily 2 types of people: 
    1. Business hunters / leads of consultancies and agencies,
    2. Management at traditional businesses (insurance, data companies).
  • It seemed like many traditional businesses are actively trying to find use cases for AI, and so when I asked them what they're doing with AI I got one of two responses:
    • It's confidential.
    • Long complicated sentences that I couldn't make simple sense of.
  • This might be naive, but it sounded to me like they're forcing AI use cases or trying to shoe horn it into their businesses, rather than having a solid business use case for it.
  • I live in a small ecosystem at Big Tech. I don't get to interact with the kind of people at the conference - people who are entrepreneurial and looking out to make professional connections and find business.
  • People were impressed by my credentials. 6 years at Big Tech made an impression.
My conversations
  • My premise when speaking with people was: I've been working at Big Tech for 6 years now, and I want to jump ship to get ahead of the AI curve. Can you connect me with someone that might be interested in me. I got some leads. 
  • The advice / takeaways were:
    • Take the plunge. You can always go back to Big Tech.
    • Take your time before you switch. You're in the position of advantage. Be super picky.
    • It takes meeting 100s of people before you find 1 person that's a good match for what you're looking for. It's a lot like finding love.
    • If you start an agency or consultancy, build a niche. Solve very specific problems because it's hard to continuously find work if you're a general agency.
    • Start something of my own. Investors in Europe will jump at me because of my Big Tech credentials.
    • Build a GPT wrapper for an old slow moving industry and sell it to them.
    • Join a medium sized company at a senior role and keep hopping to gain experience until I get referrals for C level positions.
    • Use lu.ma for finding events.
The conference
  • Interestingly, the conference is an initiative of a single organizer - Adam Fulham. What I noticed works for him is that he has got the basics right:
    • The name: It gives the ring of an established setup. It's clear what it's about, and it's got the year at the end of it which sounds like it's a yearly event that's been running for a while.
    • He has hosted these in pretty much every European city, and he has got a simple website with a clear line up of who's speaking and where and when the events are happening.
    • The price at Eur 75-150. I spoke with the guy later and he said he set the price at Eur 30 in the beginning, and the with increasing demand, he raised it to Eur 200 towards the end. The quality of credentials of the people at the tail end was a lot higher.
  • Networking is far more interesting than the panel discussions to me. Unless super technical, panel discussions are often a rehash of talks and q&a you can find on Youtube. 
  • Color coding whose looking for what by wearing a sticker would have helped so much. Else you're just meeting people at random.
  • Next time, before I sign up for a conference, I'll ask for a list of people that are coming and set up time with each person up front.

Comments

Post a Comment

Popular posts from this blog

Work Week: Sprints without the Jargon

Our team of 3 at big tech was constantly bombarded with the next big fire to put out on a weekly basis, requiring us to shift focus as soon as something came up. In the last 2 months, I've used the following strategy on  how we work , and we've found this very effective. No fancy tools, no agile sprint jargon, just the barebones of what made sense to us. Once you've figured out what the team needs to roughly do in the next year, then it's all about execution. Here's what our week looks like. Monday Morning = Plan We meet on Monday morning at a time most convenient to all and decide the agenda for the week. Each team member takes about 5-10 minutes. A minute on the past week Is there something that came up because of which you couldn't complete the planned work? Agenda for the next week Was there something that came up during the past week that you noted important enough to work on immediately? What's your focus for this week? Aim Plan only as much as you ca...
Continue reading »

AI: Asteroid Incoming

Image
Before we know it, everything around us is going to change. Here are my notes as of Jan 2025: We live in a world where computers can communicate more naturally, professionally, and even compassionately than most people. Computers have become sorcerers of words, and yet everyone is going about their lives as if nothing has happened. It feels like AI is that Earth shattering asteroid headed straight towards us from the movie Don't Look Up. If you haven't see the movie, please watch it. It’s a brilliant take on human nature. Half the people don’t believe AI will change anything. A few are sounding the alarm. And politicians? They’re doing what they do best—grabbing more power. Unfortunately we live in a hyper world, and people are tired of hypes. In 5 years, we've had Web3 Crypto NFTs AI may be hype, but for good reason. There's substance here. Of course, Rule 34 of the Internet prevails. I find it amusing that there aren't killer AI applications yet, except search and...
Continue reading »

P(ai)r Programming

Over the last 2 months with Claude Pro I've reached the point where I work with AI the way I work with humans. I'm working on a fairly non-standard app - it's got a long living foreground service and interacts with the device from that service. And I'm surprised how helpful it has been for this non-standard use case. Simple tasks First, for simple tasks, I tell the AI what to do and ask it explicitly not to write code. We first review what it's going to do before letting it code. It takes us about 2 to 3 iterations, and then it codes. For most tasks, it's really good. The reviewing started as a way to work around Claude's tight token limit, but now it's just a really good tool to make sure Claude and I are on the same page. This is exactly what I do at work with my team. Every meeting is us writing high level bullets of what they will work on, and once we're on the same page, I just review their code - which usually takes them 1-2 days to come back w...
Continue reading »