Hebb Rule is Enough - Jayan Iyer solved AGI

Hi, I have solved AGI. Impact maximization via hebbian associative learning is the best approach for AGI. plugging inputs and outputs randomly to a large hebbian spiking NN as below gives AGI. Please see the below whitepaper for why so #AGI #singularity superintelligencetoken.wordpress.com/2021/11/13/our…

@TheAusToday @VictoriaPolice you sissies are not gonna bother I presume

Shocking video emerges from alleged racist attack on elderly Indian Australians in Melbourne’s west The Australia Today understands this footage is linked to an incident in which elderly Indian Australians were allegedly abused and harassed during their morning walk, before the confrontation turned violent. This is disturbing, unacceptable and must be taken seriously. Detailed news coming on @TheAusToday #Melbourne #Victoria #IndianCommunity #Racism #StopHate #CommunitySafety #Justice #TheAustraliaToday @JacintaAllanMP @MathewHilakari @VictoriaPolice @evmulholland @MoiraDeemingMP @DrAmitSarwal @Pallavi_Aus @JitarthJai

Just wanna reiterate - Every single thing that exists, does something different from it's immediate surroundings. If they didn't, they won't be identified as a seperate thing. This is the where novelty as the fundamental starts.

@PeterDiamandis Give him a body now

You can now upload the brain of a fruit fly online. It’s NOT AI. This is a living being…online.

@tekbog Ive solved it. R U intelligent enough to realise it?

so what happened to building AGI? lmao

Unsupervised Hebbian SNN is AGI

@svembu We have to solve AGI for that. I have already solved AGI, pay attention and help promote me before others steal the credit.

My single point agenda in life is to make Bharat a technologically strong and self reliant nation, which I see as critical to our economic prosperity as well as our rural and civilisational revival. Our civilisation has to show the world through living example how to combine prosperity with humility and contentment, and show the sustainable path to harmonize humanity and the natural world. Our past 1000 year history has left us deep civilisational wounds and the defeatist, colonial mindset is one of them. Being world-class in technology is the key to restoring true pride in ourselves as a nation, and building that technology with rural talent spreads the wealth across society. We need people to do deep work in R&D. I have some skill in this area and that is the best use of my life for our nation. I view politics as public service and as a democratic duty in which citizens must be involved but given my overriding priority in life, I cannot be and will not be in active electoral politics. Bharat Mata ki Jai 🙏

@svembu Excuse me, I have solved AGI. Keep ignoring me while talking about being patriotic and supporting indian researchers.

Landed in Chennai and decided on a lark to go to the Thuglak Annual Day (aandu vizha) and asked Google search "Thuglak aandu vizha 2026 location" and the AI interjected and told me it was the Music Academy, Chennai. Meanwhile, my smarter colleague had called around (he won't trust AI on these matters!) and found out it was Narada Gana Sabha. I told the AI "wrong info, check again" and this time it comes back, corrects itself and gives me Narada Gana Sabha. Apparently it is usually at the Music Academy but changed this year. N = 1 is a hard problem for statistics 🤷

Key Points - Hebb Rule is Enough: 1. AGI(/ACI) needs to create, iow generate thought that can ultimately maximises higher order novelty on the real world. Imagine the difference between a place with intelligence and one without. See blog for what is higher order novelty. 2. Unsupervised Hebbian SNN reservoir already has the instinct to do this - it suspends output action upon perceiving novel inputs and hastens output action upon seeing the same inputs repeatedly as shown and explained in prototype 1 - doom scroller 3. Association can bring about any thought/ any kind of intelligence - either planning, reasoning, predicting, calculating , self-awareness etc etc. And a hebbian weight modification SNN already associates percepts temporally and spatially. 4. Survival/self-preservation, movement, contol, language and all such desired abilities to maximise creation emerge automatically in just an unsupervised hebbian SNN as demostrated in prototype 2 (maze runner). More helpful inputs, rules from neuroscience will make it more sophisticated. Bottom line is we dont need supervision or objective function of any kind. All abilities emerge 5. We can start with a trillion neurons on day 1 and dont need to wait for millions of years of evolution/learning. While im not saying hebbian rule is all you need, im saying hebb rule is enough to start building systems and more findings from neuroscience can be incorporated later either regarding spike timing or neurogenesis or any other hoeostatic processes - and an approximation of these fuctions can be brought even if not the exact bio-chemical phenomenon. 6. Intelligence has to create a thought - whether that thought is creative, calculative , articulate or whatever else is secondary. General intelligence is not based on the outcome but based on the amount of new processes it can generate and apply to a set of percepts towards a goal. Thngs like LLMs apply the same steps to all percepts and hence not generally intelligent. While random process can create at times too, an intelligent system is able to introduce processes that can speed up/increase efficiency of creation. I know chollet et all have said things similar to this and about creation etc, but i have been saying these for more than 5 years - just that people havent noticed. For detailed explanations if each of the above points see superintelligencetoken.wordpress.com

@yacineMTB By money, you mean followers...

I need to make as much money as possible as quickly as possible or my kid is completely done for

Trial and error infact can bring about this, but if something can beat trial and error efficiency its AGI. In any case AGI still has no goals on its own here - just applies new procedures to the goals given to it. A Superintelligence on the other hand needs to have its own goals.

While people are struggling to define general intelligence, let me point out once again that what we mean by general intelligence is not based on the outcome, but by the amount of new processing procedures it can create and bring in processing percepts towards a goal.

@QuasiCryptoOG @haider1 Painful to see hypers hype

@haider1 Not even close

sam pointed out in a recent podcast that "AGI might have already whooshed by", makes complete sense if you showed people in 2015 what AI can do today, i think most of them (even a lot of experts) would say, "yeah, that's AGI" but now it seems like the goalposts keep shifting

@iruletheworldmo 2025 felt nothing

we are about to step into the singularity. 2026 will feel 10x quicker than 2025.

@BtcBlackthorne @apples_jimmy It's painful watching believers believe

@apples_jimmy It's painful watching copers cope.

People still live in denial. They crave for a podcaster or a lesswrong post “ actually scaling is hitting a wall “ The massive compute hasn’t even come online yet, we live in very exciting times and now we move to automated ai research. To superintelligence and more.

@JannaLevin Eii

Consciousness is a strange gift, given the level of deception involved in the experience. Most everything we viscerally perceive is an embellished illusion draped over the bare facts.

@kimmonismus @GoogleDeepMind This guy knows nothing

.@GoogleDeepMind : Minimal AGI could be here in 2 years. That's when AI can do cognitive tasks without surprising failures! Full AGI means matching the full spectrum of human cognition. ASI goes beyond human limits. Lets go!

@ShaneLegg @FryRsquared Shoosh

I’ve publicly held the same prediction since 2009: there’s a 50% chance we’ll see #AGI by 2028. I sat down with @FryRsquared to discuss why I haven’t changed my mind, and how we need to prepare before we get there. Catch the full conversation here: youtu.be/l3u_FAv33G0

@rand_longevity Who killed the word proto-AGI ? That's a true milestone... Not the one that fits bigtech narrative

so did we get AGI or what?

@leecronin On top of that, doing those things doesn't make them truely intelligent. A random word generator can do those too.

Probabilistic slop engines cannot do science, drug discovery, materials discovery, or magic. Anyone who thinks AI can autonomously do science simply doesn’t understand how knowledge is created.

@xriskology Hopefully this happens. So many of us are so annoyed with those LLM people

I agree, because the AI bubble will pop, the economy will collapse into recession, and the evangelists will finally come around to Gary Marcus' view that LLMs have been wildly oversold.

@alexwg What's going on with options maxi?