Hacker News: cantor_S_drug

New comment by cantor_S_drug in "Pakistani newspaper mistakenly prints AI prompt with the article"

cantor_S_drug — Wed, 12 Nov 2025 14:04:37 +0000

Not just that we know from heavy reddit posters that they have branching universe templates for all eventualities, so that they are "ready" whatever the outcome.

New comment by cantor_S_drug in "Yt-dlp: External JavaScript runtime now required for full YouTube support"

cantor_S_drug — Wed, 12 Nov 2025 14:00:58 +0000

You are a digital hoarder. I have taken so many pics that I wouldn't even bother to look back that them (do we ever?) but Google memories is really a neat feature, it refreshes memories. I think you should run a similar service to refresh memory of your favourite videos like they are on speed dail.

New comment by cantor_S_drug in "Yann LeCun to depart Meta and launch AI startup focused on 'world models'"

cantor_S_drug — Wed, 12 Nov 2025 13:55:09 +0000

Old story : killing the goose who lays golden eggs. We humans never learn, don't we?

New comment by cantor_S_drug in "Yann LeCun to depart Meta and launch AI startup focused on 'world models'"

cantor_S_drug — Wed, 12 Nov 2025 13:51:09 +0000

Learning Algorithm Of Biological Networks

https://www.youtube.com/watch?v=l-OLgbdZ3kk

In this video we explore Predictive Coding – a biologically plausible alternative to the backpropagation algorithm, deriving it from first principles.

Predictive coding and Hebbian learning are interconnected learning mechanisms where Hebbian learning rules are used to implement the brain's predictive coding framework. Predictive coding models the brain as a hierarchical system that minimizes prediction errors by sending top-down predictions and bottom-up error signals, while Hebbian learning, often simplified as "neurons that fire together, wire together," provides a biologically plausible way to update the network's weights to improve predictions over time.

New comment by cantor_S_drug in "Yann LeCun to depart Meta and launch AI startup focused on 'world models'"

cantor_S_drug — Wed, 12 Nov 2025 13:45:34 +0000

Here's my prediction : The rapid progress of AI will make money as an accounting practice irrelevant. Take the concept of "Future is already here but unevenly distributed." When we will have true abundance, what the elites will target is the convex hull of progress, they want to be in control of leading edge / leading wavefront and its direction and who has access to resources and decision making. In such a scenario of abundance, populace will have access to iPhone 50 but the Elites will have access to iPhone 500. i.e. uneven distribution. Elites would like to directly control which resource gets allocated to which projects. Elon is already doing that with his immense clout. This implies we would have a sort of multidimensional resource based economy.

New comment by cantor_S_drug in "Yann LeCun to depart Meta and launch AI startup focused on 'world models'"

cantor_S_drug — Wed, 12 Nov 2025 13:32:31 +0000

Intelligence and human health can't be defined neatly. They are what we call suitcase words. If there exists a physiological tradeoff between medical research about whether to live till 500 years or to be able to lift 1000kg when a person is in youth, those are different dimensions / directions across we can make progress. Same happens for intelligence. I think we are on right track.

New comment by cantor_S_drug in "Yann LeCun to depart Meta and launch AI startup focused on 'world models'"

cantor_S_drug — Wed, 12 Nov 2025 13:29:04 +0000

Because when the recipe is open and public, the product's success depends on Distribution (which has been cornered by MS, Google, Apple). This is good for the ecosystem but not sure how those particular VCs will get exits.

New comment by cantor_S_drug in "Contributing to open-source should be required, like jury duty"

cantor_S_drug — Tue, 11 Nov 2025 18:06:53 +0000

Governments should set aside some pool of money for distributing to people maintaining critical projects. We have so much surplus and some portion of it should be allocated to such projects as digital infrastructure. Google did that with Summer of Code. Shareholder value can go and take a jump into a nearest lake.

New comment by cantor_S_drug in "The kind of company I want to be a part of"

cantor_S_drug — Tue, 11 Nov 2025 17:40:56 +0000

I think there was an article on English stackexchange where linguists proved that the word "the" is a useless word. It's a filler word like umm, etc.

New comment by cantor_S_drug in "iPhone Pocket"

cantor_S_drug — Tue, 11 Nov 2025 15:21:00 +0000

Can Apple lock-in those people who definitely want small phones by some prepaid arrangements which the users can't back out? That would be market working. Is there a reason why they don't do this?

New comment by cantor_S_drug in "Meta projected 10% of 2024 revenue came from scams"

cantor_S_drug — Fri, 07 Nov 2025 16:13:34 +0000

Zuck and Meta are playing the role of Robinhood. They take the money from scammers and advance their philanthrophy.

https://x.com/a16z/status/1986486508355002584

Priscilla Chan and Mark Zuckerberg on Curing All Disease

We sat down with Dr. Priscilla Chan and Mark Zuckerberg, co-founders of the Chan Zuckerberg Initiative, to discuss their ambitious plan to cure, prevent, and manage all disease by the end of the century.

New comment by cantor_S_drug in "You should write an agent"

cantor_S_drug — Fri, 07 Nov 2025 08:42:44 +0000

There is context caching in many models. It is less expensive if you enable that.

New comment by cantor_S_drug in "You should write an agent"

cantor_S_drug — Fri, 07 Nov 2025 08:39:15 +0000

https://x.com/rerundotio/status/1968806896959402144

This is a use of Rerun that I haven't seen before!

This is pretty fascinating!!!

Typically people use Rerun to visualize robotics data - if I'm following along correctly... what's fascinating here is that Adam for his master's thesis is using Rerun to visualize Agent (like ... software / LLM Agent) state.

Interesting use of Rerun!

https://github.com/gustofied/P2Engine

New comment by cantor_S_drug in "You should write an agent"

cantor_S_drug — Fri, 07 Nov 2025 08:28:47 +0000

https://github.com/mem0ai/mem0?tab=readme-ov-file

Is this useful for you?

New comment by cantor_S_drug in "Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model"

cantor_S_drug — Thu, 06 Nov 2025 20:07:18 +0000

If "not even wrong" is more wrong than wrong, then is 'not even right" more right than right.

To answer you directly, a smaller SOTA reasoning model with a table of facts can rederive relationships given more time than a bigger model which encoded those relationships implicitly.

New comment by cantor_S_drug in "Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model"

cantor_S_drug — Thu, 06 Nov 2025 18:28:34 +0000

In CS algorithms, we have space vs time tradeoffs.

In LLMs, we will have bigger weights vs test-time compute tradeoffs. A smaller model can get "there" but it will take longer.

New comment by cantor_S_drug in "Mathematical exploration and discovery at scale"

cantor_S_drug — Thu, 06 Nov 2025 17:43:30 +0000

Then its utility as a best search agent is even more. It proves the statement, LLM will find the needle in the haystack if the needle exists.

New comment by cantor_S_drug in "The trust collapse: Infinite AI content is awful"

cantor_S_drug — Thu, 06 Nov 2025 15:00:28 +0000

Neil deGrasse Tyson said a quote expressing a concern about the future impact of AI on information credibility.

The exact quote is: "I foresee the day where AI become so good at making a deep fake that the people who believed fake news as true will no longer think their fake news is true because they'll think their fake news was faked by AI."

New comment by cantor_S_drug in "The trust collapse: Infinite AI content is awful"

cantor_S_drug — Thu, 06 Nov 2025 14:54:21 +0000

We need PageRank like algorithm for "Trust / Human Content" to be applied directly to the source of such content. E.g. following all three channels are AI made. But all these content can be liked to an advanced AI version of audio based videos of Wikiarticles. If a video is providing just a summary based on established historical facts, even though it is AI based, how is it different than refering a thesaurus or dictionary? Aren't such videos making "knowledge" accessible.

FINAL Financial hours of U.S.A. just before the 1929 crash

https://www.youtube.com/watch?v=dxiSOlvKUlA&t=1008s

The Volcker Shock: When the Fed Broke the Economy to Save the Dollar (1980)

https://www.youtube.com/watch?v=cTvgL2XtHsw

How Inflation Makes the Rich Richer

https://www.youtube.com/watch?v=WDnlYQsbQ_c

New comment by cantor_S_drug in "First recording of a dying human brain shows waves similar to memory flashbacks (2022)"

cantor_S_drug — Mon, 03 Nov 2025 10:08:45 +0000

Ramachandran, the Temporal Lobes Epilepsy and God - Part 1

https://youtu.be/qIiIsDIkDtg?si=bIjpz5mWHEbN_NDI