Rinat Menyashev retweeted
- How to deal with russian colleagues?
- Skip all the good; go directly to what’s broken. Nobody gives a shit about what’s working, only what’s broken.
Personal coaches' bread and butter (and it works!)
I can’t find the original @tylercowen reference, but he said once that the one of best thing you can do for your neighbors is to raise their level of aspiration, i.e., to merely suggest that they can do something more than they thought they can do. It is *a very good thing*. Giving someone license to act (and sometimes the means), followed by action, results in success in an abnormal fraction of cases. I vaguely recall @paulg may have said that too.
A sibling of this concept is that, at an individual level, it helps to be a bit more ambitious than what your closest neighbors suggest you should be. This set includes your teachers, your manager, your friends, sometimes your parents. The suggestion by anyone that “you can’t do that” should be a call to arms.
Many Europeans succumb to low self-expectations, and the ones you find in the States often self-selected against this attitude.
1
35
Rinat Menyashev retweeted
You need to avoid Twitter for a couple of days because of The Boys series finale spoilers, or else you would know that Homelander has also joined Anthropic.
I spent 100 hours over the past week researching, writing and editing the piece we just put out.
It’s a scenario, not a prediction like most of our work. But it was rigorously constructed, dismissing it outright requires the kind of intellectual laziness that tends to get expensive.
And we’ve released it for free. Hopefully you enjoy it.
citriniresearch.com/p/2028gi…
We use this approach all the time — it’s more efficient to test the problem directly than to try to anticipate every scenario in advance
One of my favorite lessons I’ve learnt from working with smart people:
Action produces information. If you’re unsure of what to do, just do anything, even if it’s the wrong thing. This will give you information about what you should actually be doing.
Sounds simple on the surface - the hard part is making it part of your every day working process.
1
102
Rinat Menyashev retweeted
Кац объявил о создании «Фонда борьбы с фондом борьбы с коррупцией»
panorama.wtf/25236
Rinat Menyashev retweeted
The #NobelPrizeinPhysics2024 for Hopfield & Hinton rewards plagiarism and incorrect attribution in computer science. It's mostly about Amari's "Hopfield network" and the "Boltzmann Machine."
1. The Lenz-Ising recurrent architecture with neuron-like elements was published in 1925 [L20][I24][I25]. In 1972, Shun-Ichi Amari made it adaptive such that it could learn to associate input patterns with output patterns by changing its connection weights [AMH1]. However, Amari is only briefly cited in the "Scientific Background to the Nobel Prize in Physics 2024." Unfortunately, Amari's net was later called the "Hopfield network." Hopfield republished it 10 years later [AMH2], without citing Amari, not even in later papers.
2. The related Boltzmann Machine paper by Ackley, Hinton, and Sejnowski (1985) [BM] was about learning internal representations in hidden units of neural networks (NNs) [S20]. It didn't cite the first working algorithm for deep learning of internal representations by Ivakhnenko & Lapa (Ukraine, 1965)[DEEP1-2][HIN]. It didn't cite Amari's separate work (1967-68)[GD1-2] on learning internal representations in deep NNs end-to-end through stochastic gradient descent (SGD). Not even the later surveys by the authors [S20][DL3][DLP] nor the "Scientific Background to the Nobel Prize in Physics 2024" mention these origins of deep learning. ([BM] also did not cite relevant prior work by Sherrington & Kirkpatrick [SK75] & Glauber [G63].)
3. The Nobel Committee also lauds Hinton et al.'s 2006 method for layer-wise pretraining of deep NNs (2006) [UN4]. However, this work neither cited the original layer-wise training of deep NNs by Ivakhnenko & Lapa (1965)[DEEP1-2] nor the original work on unsupervised pretraining of deep NNs (1991) [UN0-1][DLP].
4. The "Popular information" says: “At the end of the 1960s, some discouraging theoretical results caused many researchers to suspect that these neural networks would never be of any real use." However, deep learning research was obviously alive and kicking in the 1960s-70s, especially outside of the Anglosphere [DEEP1-2][GD1-3][CNN1][DL1-2][DLP][DLH].
5. Many additional cases of plagiarism and incorrect attribution can be found in the following reference [DLP], which also contains the other references above. One can start with Sec. 3:
[DLP] J. Schmidhuber (2023). How 3 Turing awardees republished key methods and ideas whose creators they failed to credit. Technical Report IDSIA-23-23, Swiss AI Lab IDSIA, 14 Dec 2023. people.idsia.ch/~juergen/ai-…
See also the following reference [DLH] for a history of the field:
[DLH] J. Schmidhuber (2022). Annotated History of Modern AI and Deep Learning. Technical Report IDSIA-22-22, IDSIA, Lugano, Switzerland, 2022. Preprint arXiv:2212.11279. people.idsia.ch/~juergen/dee… (This extends the 2015 award-winning survey people.idsia.ch/~juergen/dee…)
Rinat Menyashev retweeted
the new o1 model looks amazing but luckily it has a phd level intelligence so our jobs are safe for now
Rinat Menyashev retweeted
Quantum CEO explains how his startup intends to scale from 5 qubits to 5 million qubits.
didn't know about a double descent phenomenon, looks like magic to me
Just want to reiterate that LLMs are magic (Claude Opus specifically)
Here's what I got done in 5 mins:
• Was watching @ylecun explain double descent youtube.com/watch?v=5I7-ItaO…
• He suggested that you can observe the same phenomena with polynomial models too
• Asked Claude to write code to help visualize this
• Fire up Jupyter notebook and the code works flawlessly in one go
Lo and behold, what @ylecun said appears clearly in the plot!
This is as close to magic as you can get today!
139
Rinat Menyashev retweeted
"If I Had More Time, I Would Have Written a Shorter Letter" -- Blaise Pascal



























