Keith Patarroyo

I’m certain that there is something to it, a lot of my thoughts for the last 4 years are merged there. But I still need to properly write it and check if things actually hold together. I have a hunch that it wasn’t a fever dream…

I thought strongly about novelty search and how to apply it to my current work in open endedness in cellular automata. After a lot of good ideas I constructed a proof of the possibility of infinite scalability in the universe. Just like Poincaré I wrote it crystallised in a piece of paper.

Yesterday I had a very similar experience to Poincaré. I wasn’t able to sleep because I drunk a coffee at night and I was focused for several hours to synthesise my thoughts about evolution as a fundamental law of the universe. It was a coincidence because I read what I needed before going to sleep.

Whenever a dynamical law divides the state-space into such separate cycles, there is a memory of which cycle they started in. Such a memory is called a conservation lan, it tells us that

I’ve been casual student of physics for four decades. Why did no one ever explain to me so clearly what a conservation law is?

From Susskind and Hrabovsky (2014) The Theoretical Minimum.

—

This is why reading basic material from brilliant thinkers is so often worth the investment.

Catholicism is fundamentally pure suffering.

Of course there are exceptions to all of these, however the more the person is immersed in the system, the less of an exception they will be. Also, there are outliers, but honestly, probably none of this applies to them, so who cares. A truly remarkable individual will make his own journey anyways…

The reason for this phenomenon is that probably the 30something has something missing and that something is not easy to fix. Whereas the 20something has a higher chance of having that something easy to fix. It is a bit heartbreaking, but the world is harsh.

At your 20s you are mostly potential and people will take risk by taking you, but at 30 if you have not shown any accomplishment, the risk is a lot larger, and the return is unknown. Compare this to a similar profile with modest accomplishments or even a 20something with your exact same profile.

As a minimum, you gain discipline and work ethic, which is essential in the future. By the time you hit 30, things get a lot more murky, because you have enough data points in your CV for people to figure out your capabilities. And this is why people will treat you differently from where you are 20…

Unless you choose a very fixed career like medicine or law, things are very open ended. The only advice I have is that you should choose something, anything really, and pursue it. Not choosing is a terrible idea. It is better to choose wrong, and change later rather than not to choose at all.

Yes, this IS the difficult part of the journey… the exiting part is the risk! To give structure to the risk is the difficulty. Honestly, I don’t think there is a right way. It is all very much up to chance, where do you start, to what do you build upon.

It’s easier to focus in your late 20s, but by then you should have a foundation of where to pull. At this point things sound quite harsh, you don’t know where to start, and you are supposed by mid 20s to build on something, and by 30 you should have a point of achievement(I’ll come to this later).

In general generating routine by yourself is a bad idea. Whereas it’s easy to take risk by yourself. The question of what to focus on is hard, this is what you will probably struggle with in your early 20s. That’s why college is an easy option, because in choosing a major, the options are very wide.

People are not in the mood to train a 30 year old from scratch. However focusing too much is a danger, if you don’t take any risk, then you played safe the time of your life where you could take the most risk. There is a balance and everyone has different weaknesses here.

At the same time you should focus on something, this should generate some sort of routine and you should acquire ability from it. College is the default option here, but it can also be an apprenticeship or just regular work. This is important, because people expect by 30 that you can do something.

I can say that this time is incredibly exciting and difficult. In general it is a time where there are few expectations of you and you are allowed to take a lot of risk. It is not that you are not allowed to take risk later, but you will have more to lose and society will be much less forgiving.

I can mostly talk about the ages I’ve experienced and the ones that I’m close to. That would be early 20s, late 20s and 30s. While I’m not 30 yet I can talk about certain elements that ring true. So I’ll start with general comments on the 20s.

Some people have called me ageist in the past, however if anything I’ve become more ageist as I get older. While I’m not making definitive statements about age, or about things that happen at a specific age. There are certain elements that can’t be denied.

It’s incredible to have witnessed two technological revolutions in my lifetime. It is not clear to me that a third one will occur in the early 2030s. Also it will likely not come from Google or OpenAI… The future and my future is still open…

second AI revolution. Similarly to the one of the previous decade, people will only start to catch up around 2026-28. Until then it will be an amazing time for the user of this technologies, and after that the real bubble will start to crack...

Even if I would‘ve been in the correct path my general opinions on AI around 2018-22 were way to heterodoxical. I would have likely focused on an aspect of AI that interested me, but unlikely to have any impact on the current foundation models. All of this support my hypothesis that we are in a…

Because I did my graduate work 10 years after Ilya did. But our results were very different… and even If I would have shifted earlier to AI, like my best friend did in 2016. It wasn’t clear where the golden goose was, unless you got a OpenAI internship around 2018…

It’s also quite surprising how AlexNet and ChatGPT are 10 years apart. Also makes me think about the role of luck in all of this… I’m pretty much 10 years apart with Ilya, but if I would have done my PhD in NLP, I would’ve been become obsolete. It is really hard to compare…

Something very similar may have happened to the previous generation of Deep Learning, I.e 2008-2012. People saw how neural nets solved: image, speech recognition, translation, recommendation systems. This should have changed a lot of people minds… specially if they worked with rule based systems.

The Keith of 2018-2022 would be appalled by how mainstream my opinion on LLMs and AI has become… The capabilities of LLMs completely changed my mind, I used to argue for alternative architectures, new math, new algorithms, a complete change of framework… but no longer, it may have to do with age…

Sailed for the first time, little wind… not suuuper fun, but I’m amazed, it’s one of the few activities with an unbalanced female/male ratio. I think it was roughly 3/1. Maybe a complete fluke…

Vacations are over, very mixed feelings when telling people how things went… It will take time to process everything by myself. I mean, almost everything was like a dream…

Update #9.5: Grape picking in a vineyard in Tuscany. The ultimate dream… unless you have a meltdown… Vacations started badly and ended terribly, this seems to be a pattern… This will be one very difficult to process, because I feel I was wounded quite badly…

Update #9.0: Saturday Market, Tarot reading of the future, lunch at a terrace overlooking the valle and sunset in a stunning location. Somehow under the sunset light everything shows its true nature... I just never expected it was going to be so ominous… the message was grape danger…

Update #8: Eating even more fish, improv in our terrace, eating under the moonlight, movie night, stargazing in a mountain, nudist beach, chanting session and a lot of fun. I’m so happy spending time here with my friends that I didn’t feel compelled to travel anywhere else… It’s been that fun!!!