bobosha

off-topic: i've never seen a paper with this many authors.:-)

Our time there overlapped—I was a student of TimBL from 2004 to 2006, researching the Semantic Web initiative. CSAIL was, and still is, a truly magical place. I had the privilege of working alongside some of the greatest minds of our time: David Karger, Manolis Kellis, Regina Barzilay, et al an incredible assembly of brilliance.

i hired 2 engineers in Argentina through that thread and they were some of our best resources.

Mining visual data—whether from CCTV, smartphone videos, or other sources—is the next big frontier. Given its immense information density, the potential payoffs could far exceed those of previous data revolutions.

This is also common in the Italian/Greek/Indian etc. American subcultures. Immigrant cultures take their time - and sometimes never - in absorbing the larger anglo-saxon culture into their own.

>My only consolation is knowing that I'm seeing the world more accurately than I used to.

also known as growing older ;-)

The rise of islam was only possible due to the fight to exhaustion by the Byzantines and Sassanids. If not for the timing, Muhamad and his religion would have been but an obscure cult in the sands of Arabia. Just goes to show that timing is everything in history.

Streisand effect in full display.

Rosy Retrospective Bias [1]

[1] https://en.wikipedia.org/wiki/Rosy_retrospection

A better system would be a points based one: that uses salary indexed to COL, educational qualifications etc.

well said, this debate is too emotional to have a balanced conversation. Overall h1b issue is a dichotomy - like with many such hot-button issues - with both sides right about the other side, but wrong about themselves.

tyvm for your thoughtful answer without some of the usual histrionics on this topic.

The results look quite promising.i will give this a try...

I had the privilege of spending 1.5 years (2004-06) at MIT CSAIL under his guidance, where my research focused on the Semantic Web—an idea aimed at making the web more intelligent. With that experience in mind, let me share a couple of thoughts:

1. Even after all this time, I believe that the vision of a web of data and knowledge at a global scale remains incredibly compelling. It may still represent the next major evolution of TimBL’s original invention.

2. However, it's increasingly clear that this leap forward will likely be driven by machine learning rather than traditional methods like RDF/OWL.

Actually, no, I didn’t have a warm introduction. I reached out to him directly and got a response within 5 mins!, I was inspired by another entrepreneur who successfully raised funds in a similar way. (By the way, his email is publicly available and easy to find online.)

Several factors worked in our favor:

- We had an MVP and paying clients.

- In 2017, AI and machine vision were gaining traction and generating a lot of buzz.

- Plus, my academic background likely helped as well.

In 2017, I secured investment from Mark Cuban for my startup through a cold email. He contributed to our initial seed round, followed on in subsequent funding rounds, and we successfully exited in 2023.

My email was concise, under five sentences, and straight to the point, highlighting the market problem, its size, and our traction. Feel free to ask any questions!

I'm developing an implementation of what I call Hydra – Multi-Head Prediction Embeddings [1], which I believe represents the next evolution in transformer architectures.

[1] https://www.researchgate.net/publication/381009719_Hydra_Enh...

am i the only relieved that for once a non-AI human has won the nobel?

I am a proud Indian-American (naturalized after h1b) and I think this kind of discriminatory mindset has no place anywhere, definitely not in the workplace. I think my community can be secure in its success and learn to do better and be open to critique. Let's work to abolish these attitudes by being more self-aware and raising awareness & use cautionary tales like this as a wake-up call.

Schmidhuber invented everything. /s

haha!

i have the book and tried to read it many times, but never grokked it tbh. ;-)

Could you give us a tl;dr on this project? and how could I use something like this work for on-device applications, think "smart home" style applications?

In my experience, "founder mode" is when founders intuitively understand what is best for their business and steer clear of cargo cult practices. All too often, professional management falls back on formulaic methods that end up driving businesses into the ground. This post [1] is a great case study of one such instance.

[1] https://healthio.notion.site/How-Cargo-Cult-Thinking-Nearly-...

Although the author points out that there's been a "huge amount of writing" on the topic, the issue persists due to the myth of "I have an MBA, therefore I know better." MBA programs often instill in their graduates a misleading confidence in their management abilities, dare-I-say smug arrogance, which, in my experience, frequently leads to the implementation of cargo cult practices [1] with potentially disastrous results.

[1] https://healthio.notion.site/How-Cargo-Cult-Thinking-Nearly-...

and some of us find it does a lot more...I have been increasingly relying on ai-generated code...it's been remarkably effective in automating nearly 90%+ of what-i-now-know-to-be tedious, manual tasks.

Perhaps a good way of thinking about this is: "history doesn't repeat itself, but it rhymes" - (supposedly) Mark Twain

> It seems like people of European ethnicities are the only people on earth who don't practice this brazen ethnic nepotism

I beg to differ, Eastern European rivalries/grudges are no less: Russian v Ukrainians/Poles, Poles vs Czech and so on...

One of the best advices that I have learned is "take something that people do, and remove a step (or make the step super easy)". Perhaps that provides some fresh perspective on your conundrum.

you could do the following, use a webcam :

1. build a yolo based cat detector to detect where yuri is

2. then build an action recognition on video (there are many good action recognition approaches such as videoMAE etc that could be leveraged to detect normal cat activity vs abnormal activity.

hope that helps.