mbaytas

immediately ordered the book

fascinating checkout flow

Improving model capability with more and more data is what model developers do, over months. Structure and prompting improvements can be done by the end user, today.

In fact I think this is a ripe space for opportunities to build new institutions that provide value.

The value of college is not only education, but also community.

We are seeing, and will continue to see, many new online and IRL ventures that provide community and learning to young people, which will serve them much better than colleges.

Buildspace (https://buildspace.so/) is a beautiful example.

> I think college is broken. But my question is whether it’s broken and useful or broken and harmful.

A better question IMO is: Is college more useful than the alternatives?

College is a huge investment: tens of thousands of dollars, plus 4 years of full-time work.

When I went to college in 2007 (not in US, but I digress…) it made sense to go in for a 4-year degree because there were no alternatives. Coursera, Udemy, etc. did not exist. YouTube was just born. And to start with, we didn’t even have broadband in my house or school, that could stream video.

It’s a very different world today. The opportunity cost is too great. Even without being a Thiel fellow, there are many alternatives to what you can do with 4 years of your life that will set you up for success – not only educational opportunities but also starting a business, freelancing, etc.

Exciting! A few years ago, as an academic design researcher, I pondered the same question and wrote a few research papers.

Like the author, my inspiration for doing this work had come from a vintage watch I received from my dad!

Turns out, there’s a rich literature in design and psychology research on the heirloom status of objects. And more generally, on why and how people imbue things with value.

Here’s a link to the page from my academic projects archive, with citations: https://www.baytas.net/work/computational-heirlooms

We even went as far as to apply for (and receive) a grant for UX design research, exploring how to create for heirloom value in the digital world. I’ve moved on from academia since then, but AFAIK our PhD is working on it at KTH in Stockholm: https://www.baytas.net/work/digital-preciousness

> “In the exercise, someone picks a number of adjectives from a list, choosing ones they feel describe their own personality. The subject's peers then get the same list, and each picks an equal number of adjectives that describe the subject. These adjectives are then inserted into a two-by-two grid of four cells.”

> “Room one is the part of ourselves that we and others see. Room two contains aspects that others see but we are unaware of. Room three is the private space we know but hide from others. Room four is the unconscious part of us that neither ourselves nor others see.”

I wonder how the fourth room, which neither the subject or their peers see, is populated…

Sheet brilliance.

The landing page is indeed very good design.

As an experienced web designer I can tell you the design and implementation are straightforward, but very competently done. There are many things like the typography and UX writing that suggest this is the work of a fellow experienced designer. There’s much more here than what can be captured in a “guide” - it’s just an overall high level of design competence.

I re-drew the artwork based on the originals from the publications, it’s not copy-paste images.

This, plus the fact that the original source is cited, makes it work.

A while back I made a small NFT collection with my favorite CS illustrations:

https://instagram.com/citationsnft

It was an experiment that didn’t do much commercially, but we learned a lot from it.

I have a theory that when the C-suite decides to slay an entire division they assign an incompetent boss. Give them a few months to mess it up. Then go for the kill. The incompetent new boss magically absorbs all the blame.

Copywriter.

Branding.

Member of elite institution (MIT) - check. Expert in hot field (AI) - check. Famous guests - check.

Anyone who checks these boxes can attract a large following for a podcast, youtube, etc.

MIT is perhaps the magic ingredient with the most brand power, through which he was able to attract initial famous guests. From there, snowball.

This is not to say that Lex is not hardworking, intelligent, consistent, etc. But being a MIT AI researcher is a vastly leveraged starting point.

> “Ph.D.s are looking at the labor market, seeing opportunities out there, and taking them,” she says. “Those skills that we teach our Ph.D. students are in demand.”

LoL.

Among all the stereotypes in academia perhaps the truest one is profs taking credit for what their students achieve.

Not even bothering to hide it anymore.

There is a branch of computer science, Human-Computer Interaction (HCI), where exactly the ideas in this article have been playing out for a few decades already.

Earlier this year I wrote about HCI as the intersection of technical craft and social sciences: https://www.designdisciplin.com/hci-profession/

Interestingly the responses I got from people in the HCI field was the inverse of the sentiment in this article. Here it seems that the author is arguing to increase social sciences in CS, and getting pushback. One of my arguments was to increase tech/eng competences in HCI, and I got pushback on that too.

One thing that strikes me from this article is a vast bureaucracy, regulation, credentialing structures, accounting, etc. that the story revolves around.

In Sweden, where I live, there are plenty of educational opportunities available to people completely free of charge. Entirely subsidized by the government. Instead of assembling elaborate bureaucratic structures, the state spends its money paying school admin and teachers.

While these are not universities per se, the opportunities they open up to students are top notch. You don't need an entire college education to "set graduates on a path towards breaking the cycle of poverty", which according to the article is what work colleges aspire to achieve.

This vocational education is also much more economically efficient than universities in the same country, mainly due to losing the weight of bureaucracy and credentialing structures. Teachers are well-paid, ops run smoothly with few personnel, and students get access to many new job opportunities completely free of charge.