That’s unfortunately not how business works. You need to be brave at the start, talk about things that you are not sure about and vision that is not yet properly tested. If you set up a new leg of business then you probably fake it until you get enough clients to be 100% confident about what you actually do.

I wrote recently about how I use AI and one of the key points was that I always do a proper draft for AI. We are getting stupid and distracted because of AI, but it will be an edge for every person that will be able to use it properly and maintain clear, deep, thinking skills.

7 out of 11 diamond sponsors at #ICML2025 in Vancouver are related to trading. Two lessons from that: 1. We know who is interesting in ML innovations the most. 2. Another proof that trading companies need to evolve constanty and need to be aware of it. I wonder where the fight for talents is bigger now - in AI or in trading. Three or four years ago I would definitely bet it's trading, but now who knows. #ICML2025

I talked with SBF about FTXs latency spikes exactly 2 months before FTX collapse. During those 2 months I was suprised twice: 1. First time that he was so engaged that he wrote directly to me about the latencies. 2. Second time when I realized it was all fugazi.

In most teams, 20% of people do 80% of the meaningful work. If you’re young and ambitious - sit down for a moment and ask yourself: “Who are those people in my room?” Then, next quarter, focus on helping them. Do their dirty work. Ask them real, important questions.

I wonder how many illegal strategies started with this pattern: 1. Notice the pattern in the market. 2. Use it on a small scale. 3. You reach capacity. 4. What about actually being that pattern and using it on a big scale? For example: Pump and Dump. You notice a pattern where retail is buying a too-big-for-this-marketcap amount of a coin, and you start by front-running it. But then you start thinking - what if I create this pattern myself and run it on a much bigger scale?

I wonder how many illegal strategies started with this pattern: 1.Notice the pattern in the market. 2.Use it on a small scale. reach capacity. 4.What about actually being that pattern and using it on a big scale? For example: Pump and Dump. You notice a pattern where retail is buying a too-big-for-this-marketcap amount of a coin, and you start by front-running it. But then you start thinking - what if I create this pattern myself and run it on a much bigger scale?

Recently, my wife asked me „What do you regret in your life?”. I told her that the only thing that I regret is that Gerko and Schlamp are not on @X.

That's how it works in most of the very, very competitive disciplines. You just simply maximizing PnL with constraints regarding the law and reputation. Formula 1 is a good example - you always have a philosophy of doing everything to be the best but within technical regulations. You accept that you are going to be checked (or sentenced0 by regulators or other advisory board and your goal is just to defend yourself.

Big sign of poor management skills is creating meetings for n-best/most important people in company (n > 3) regarding a problem that should be solved with one or two persons. Complete waste of time. Such meetings are boring, most people check their emails or other messages and nobody thinks deeply about the problem. Don't do it – never. If you are a manager and have a problem: 1. Ask two people to prepare themselves for meeting. 2. Find 15-30 minutes slot 3. Discuss the most important parts 4. Give exact responsibilities. 5. Go to next problem. Regarding meetings -> "Default small, expand when evidence says so".