j7ake

This is also the philosophy espoused by Taleb. You want that job that involves tinkering, playing, and an option to select winners.

John Von Neumann would disagree with you :

"If people do not believe that mathematics is simple, it is only because they do not realize how complicated life is."

With the use of chat bots and agents, managing these on prem should be easier than ever right?

A drug is almost never acquired that can then be put directly to patients and sold.

Large pharma makes strategic bets on several drugs, some initiated in house, others acquired, but they all just go through further optimization and testing before it is approved.

Huge RnD is required even if drug is “simply acquired”

Rather than scoff at the selection, maybe use the opportunity to recalibrate and expand what one considers “interesting”?

I meant specifically the book, which doe not have any of those things you mentioned.

Also nobody fits neural networks and use variation inference using any priors that aren’t some standard form that makes algorithm easy

Is Bayesian data analysis relevant anymore in the era of foundation models and big data?

Salary range there is like 30 to 70k euros

Are there links to YouTube videos of recordings? Or do I need to find it elsewhere ?

Implicit in the funding issue is the inability to attract and retain top researchers in resource heavy fields (AI, experimental field).

The starting package For new professors at Oxbridge is several orders of magnitude less than top institutions outside UK

The universe packs in even more concepts: only 3 or 4 dimensions

Combine pass with qtpass on osx and pass on iPhone and you’re golden

There are plenty. The author is over exaggerating the prestige of CMU compared to other institutes outside of the USA.

Just in the western countries:

Toronto, Cambridge, ENS in France, the many max Planck institutes in Germany (eg Tubingen), the two federal institutes in Switzerland.

Faculty positions in any of those are likely to be better than CMU (in terms of start up package, funding, quality of students, quality of faculty, and ability to hire people).

It this lower wage true for states like California?

The totality of human knowledge is a rounding error to what’s needed for AGI

Government workers are funded by public money, so public pension monies are funded by public money ultimately.

Wait did I see “ Ontario Teachers' Pension Plan” as an investor?

Are they putting Canadian public funds into Anthropic?

Good luck trying to get a work visa to places like Europe or USA if you have a trades background

University is a path to a more global and flexible career for positions that don’t even exist yet.

There is more mobility and more flexibility to future jobs with a university degree like mathematics than with trade.

If these people are making that much, I wonder what is the compensation for people like LeCunn.

I love this podcast and substack. He is an artist in many ways : revealing from common observations into something revelatory, curious, and funny

Glossy results are memeable stories that editors of journals would like to have in their next edition.

There is very little incentive in publicly criticising a paper, there is incentive to show why others are wrong and why your new technically superior method solves the problem and finds new insights to the field.

There is zero incentive for the researcher personally to publish null results.

Null results are the foundations on which “glossy” results are produced. Researchers would be wasting time giving away their competitive advantage by publishing null results.

Maybe what’s common is that self taught people do not even know what to look for to improve on.

They only see the superficial or easy-to-spot goals and lack the eye for detail to build foundations and technique.

Like piano it’s important to tackle courses that are appropriate for your level

Self taught people often skip too much of the basics so they struggle to properly tackle the fancy stuff

Compression is fine for consuming information. But people need to be both consuming and generating knowledge in order to understand concepts beyond summary points.

Generating knowledge is an exercise in compression. It is helpful to deliver insights to readers. Consuming knowledge is important to keep your pulse in your field. Only consuming leads can create an illusion of understanding but no real usable knowledge that can provide value to the community.

The issue is that we live in an information glut and one can now live a life spending all their time consuming, but never creating for themselves.

The triple star questions in the back of exercises of textbooks will be of this calibre.

In CS check Knuths book.

“runs a lab” is putting it lightly. He’s a Max Planck institute director. This means essentially limitless funds for his own research and can shape hiring of faculty in his institute.

Chinese Buddhist cuisine is essentially the art of tasting exactly like meat without the meat.

Funding training programs and basic is extremely effective per dollar spent.

Where else can you pay someone 50k a year and get PhD-level productivity?