zmachinaz

Just zooming into the tail section photo one can find on main stream media shows very suspicious signs of possible shrapnel hit ...

The selling point is not absolute return, but uncorrelated return. If that is really the case, is a different story ...

One could equally say that mainly the mediocre people are excited, while the smart people actually see the limitations. For people doing non-trivial things, the output of ChatGPT seems currently rather far away from being of any deep value. To replace some mundane tasks, yes, but mundane tasks are anyway just for the mediocre.

Not sure that I would recommend actually hiring that guy. He does not seem to understand that modelling is only one part of the equation. If you model something non-trivial, usually you can not just hand over your R scripts to someone else and say: Please implement that in python/C. You have implementation constraints which feed back into the modelling itself, like latency or scalability. Furthermore, good luck in letting someone translate your non-trivial math into another language hoping that he won't break it in some subtle or non-subtle way. Its just far more efficient if you have someone who can actually do the prototype directly in python or C, and let then a pro-developer optimize specific parts of the code.

Do you think research equipment was cheap during the time of "aristocrat scientists" ? The point being here "after they succeeded financially".

Might go back to this model over time. Guess it is no secret that academia is in strong decline. Lot of excellent people just don't bother anymore to enter that circus. They might do their own independent research after they succeeded financially.

At its start, the quanta magazine was a good read. But, it degraded very quickly after they started to use it mainly to put scientists from under-represented groups into the spot light. Nowadays not worth any attention.

Regarding 1)

I am not sure if you are not trading "high human efficiency" against increased risk of blowing up at some point. Good luck doing forecasting without thorough understanding of priors and statistics in general.

In order to get admitted to University of Tokyo, you have to be top on the common national university entrance examination, and then be top on the University of Tokyo entrance examination. So only the best of the best in Japan gets admitted. Highly competitive, as a degree from that University sets you for life.

Problem is, there is a big divide in user base: For a normal user, 8 GB is mostly sufficient. If you have a lot of chrome tabs open, 16 GB makes your day. But, if you are any sort of a creator, you can not have enough ... my 256GB is barely enough.

Article misses the point.

By now, the good students and post-docs know what is going on: The system is already so broken that there is a high chance that you end up with a PI who got his job because he/she/* is good in making friends with some VIPs. For the actual research, that is what you are hired for. Have fun turning the "great vision" of these PIs into something that makes at least a bit of sense. And only if you are as well the type to make friends with VIPs, you will have a chance to tenure-track.

The article totally misses the point. Yes, for the digital morons (their turks) google may work perfectly fine. However, for people who actually look for real content google gets worse and worse. All the good material gets far lower ranking than the repetitive low quality medium posts with identical and non-existing content.

If you try to connect via VPN, Discord insists on a phone number for verification. That says all ...

The current version of the python lib seems to be extremely badly written code. Or is the algo so bad ? Takes something like 21s to compute the correlation for just 10k samples.

There has been work of similar spirit going on since at least 2017 in mathematical/theoretical physics. With similar limited importance. But the ideas are floating around already for a long time. But if DeepMind gives it a go they make a big fuzz about it, and of course they even do not need to cite previous works ...

I have been using ROCm for 2y+. The investment in this infrastructure was a big mistake. The biggest burner was the need to do a clean install on each new ROCm release. Clean here means manually finding and deleting all traces from the previous ROCm version, and recompilation of apps like pytorch. Good upgrades took hours, bad ones days ... . Finally I settled to freeze the system and not touch it anymore until retirement of the cards, hopefully soon.

I always thought that the top talent is actually leaving Germany to seek better opportunities elsewhere (already since 20y+). Main destinations Switzerland, UK, ...

I think its a good summary, but one should stress a key point:

All this QE, aka money printing, can only be removed in a "nice" way if the additional liquidity is to a larger extend used to create additional value rather than consumption.

De facto, most of the liquidity is consumed without added value (Zombie companies, certain government expanses,...)

This can only work for a finite amount of time.