The article describes a unifying mathematical framework for representing neural network architectures. It can be used to describe existing ones, like transformers and convolution, but can also be used to generate entirely novel ones. As part of our research, we generated several thousand new architectures, some of which show remarkable properties. The corresponding research paper is up on arXiv:
Consider the past four months of life through the filter of from?site=debarshibasak.github.io : We've gone beyond DevOps, and have suffered the Death of the Roadmap. It is acknowledged that the Easy part of building is over, and We wonder if GitHub is still needed. Many things remain unclear of course, due to the Cambrian Explosion of AI.
Disclaimer, my summary is human generated
debarshri 22 hours ago [-]
This blog is human generated too. I am just expressing my thoughts and what I am observing around me.
verdverm 19 hours ago [-]
sometimes it's best to take HN with a grain of salt
Gemini has to be the Opabinia of the AI explosion then? :)
scared_together 16 hours ago [-]
Open source models: trilobites
ChatGPT: Hallucigenia
Claude: Haikouichthys
????: Anomalocaris
Hand-written code: Ediacaran biota :(
dev213 12 hours ago [-]
Ugh, I am so tired of people drawing a comparison between large language models and life. They’re so different they’re not even comparable.
debarshri 11 hours ago [-]
I respect your perspective.
Everyone is trying to build a POV on agents, AI and LLM, how it is going impact everyone and indivually. I believe, lack of any baseline for a tech shift like this, people to draw references from theories and anecdotes to build their understanding. Thats all what I am doing.
butlike 7 hours ago [-]
I imagine the change from oil lamps to electricity was a similar paradigm shift. What do you mean one company controls if I can or can't light my home?
pvaldes 13 hours ago [-]
As AI is used to do analysis, why not Cambridgian Analytical Explosion?. Is a boundary much more clear in the Antropocene stratification, marking the start of the Robozoic.
thelastgallon 10 hours ago [-]
Cambrian Explosion of AI Slop
butlike 7 hours ago [-]
Cambrian splooge
9583sG 12 hours ago [-]
Is a beige background now mandatory for the AI cult?
This is just the daily agitprop to keep people going.
debarshri 11 hours ago [-]
The blog is not AI generated. I typed it out. The website yes. The goal was to keep it clean, I didnt care about what color it was. But then you are seeing it in daytime. It also has a night mode.
cindyllm 11 hours ago [-]
[dead]
paramahans 18 hours ago [-]
This term "Cambrian explosion of AI" was used by Jenseng Huang of Nvidia and the author could have atleast credited this to Jenseng or mentioned something about his talks or speeches
debarshri 17 hours ago [-]
I didnt know that. I'm not going to give credit to every person out there who used the term cambrian explosion.
RugnirViking 15 hours ago [-]
i'm sure jensen huang will be fine. He has one or two other accomplishments to his name
idiotsecant 11 hours ago [-]
And was famously the only person to use the phrase Cambrian explosion, ever
qarl 11 hours ago [-]
I hate to rain on your parade... but I used it too.
idiotsecant 7 hours ago [-]
Not sure if woosh or meta-meta-meta troll...
qarl 7 hours ago [-]
Yeah. I like to think of myself as the Nathan Fielder of HN.
https://www.linkedin.com/pulse/announcing-universal-mathemat...
The article describes a unifying mathematical framework for representing neural network architectures. It can be used to describe existing ones, like transformers and convolution, but can also be used to generate entirely novel ones. As part of our research, we generated several thousand new architectures, some of which show remarkable properties. The corresponding research paper is up on arXiv:
https://arxiv.org/abs/2605.04325
Disclaimer, my summary is human generated
AmazingEveryDay has a lot of critical takes here
https://news.ycombinator.com/threads?id=AmazingEveryDay
I'm not sure it was that so much as the exponential improvement in hardware over many decades reaching roughly human equivalent levels.
I mean check out Moravec's 1997 paper predicting things will get funky in the mid 2020s long before anyone had thought of GPTs https://www.jetpress.org/volume1/moravec.htm
ChatGPT: Hallucigenia
Claude: Haikouichthys
????: Anomalocaris
Hand-written code: Ediacaran biota :(
Everyone is trying to build a POV on agents, AI and LLM, how it is going impact everyone and indivually. I believe, lack of any baseline for a tech shift like this, people to draw references from theories and anecdotes to build their understanding. Thats all what I am doing.
This is just the daily agitprop to keep people going.