Home
Creators
Search Recent Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register Favorites Keys Review DMs Logout
Community
Telegram ThePornDude
Home Creators Posts Import Register Favorites Logout

AI Levels - where we are and where we are going (Patreon)

Published:
2024-07-12 22:59:11
Imported:
2026-06
Tags:
haven't archived this post yet. have a subscription? use the importer!

Content

REM THESE ARE MY VIEWS AND MY VIEWS ONLY - I could be way off!

The chart from OpenAI outlines five levels of AI development, each representing a significant leap in capabilities. Here is a quick explanation of each level and an assessment of how far we are from achieving Level 5:

Levels of AI Development

✅ Level 1: Chatbots, AI with Conversational Language DONE

  • Description: This level includes AI systems that can engage in basic conversations with users. Examples include customer service chatbots and virtual assistants like Siri and Alexa.

  • Current Status: Widely implemented and used in various industries for customer support, personal assistance, and more.

✅ Level 2: Reasoners, Human-Level Problem Solving

  • Description: AI systems at this level can solve problems at a human level, using reasoning and logic. They can understand context and provide solutions to complex issues.

  • Current Status: Advanced AI models like GPT-4 exhibit some reasoning capabilities, but they still fall short of true human-level problem-solving in all contexts.

COMING SOON Level 3: Agents, Systems That Can Take Actions

  • Description: These AI systems can perform actions autonomously based on their understanding and reasoning. They can execute tasks without human intervention.

  • Current Status: Autonomous systems, such as self-driving cars from Tesla and robotic process automation (RPA), are in development and real close to deployment IMHO

3 YEARS FROM NOW Level 4: Innovators, AI That Can Aid in Invention

  • Description: AI at this level can assist in creating new inventions and innovations. They can generate novel ideas and solutions that go beyond existing knowledge. My hope is it can cure things like cancer etc.

  • Current Status: AI is beginning to assist in research and development, such as drug discovery and material science, but human oversight and creativity are still crucial.

END GAME Level 5: Organizations, AI That Can Do the Work of an Organization

  • Description: The ultimate level where AI can manage and operate entire organizations, making strategic decisions, managing resources, and optimizing operations without human intervention.

  • Current Status: This level remains theoretical. But when it happens there will not be much for many of us humans to do.

How Far Are We from Level 5?

5 years, I would say...... given the pace of Dev. Perhaps 10 years.

Files

Previews only

Comments

Richard

Crazy to think level 5 is only 5-10 years away! Brilliant summary, thanks very much! 👍

Sher Global

Just reading Ray Kurzweil’s book ‘The Singularity is Nearer’. It seems that they think we will be merging our brains with computers in the near future. Apparently DARPA is funding something called nanograins to interface with neurons and create a ‘cortical intranet’. And apparently we’re headed for Nanoscale electrodes inserted into the brain through the bloodstream. …that is very disturbing to me on many levels! Also fascinating.

btc Bobby

Level 4 in 3 years seems crazy. But so did fsd and colonizing mars.

Sher Global

We’re in a technological exponential growth cycle that’s about to take off. ‘Moore's law is the observation that the number of transistors in an integrated circuit (IC) doubles about every two years’ This means that the computer hardware will double in processing power ever two years. Up until now that still meant that everything was slow. But we are just hitting the hockey stick part of the curve. It will double in processing power every year. We are now approaching the point where we have the processing power of the number of neurons in a brain. A computer will be able to process information as fast as we do. But that is our limit. The next year the computer capacity will double again and we will soon be left in the dust. AGI is soon. Elon predicts 2026. Futurist Ray Kurzweil thinks 2029, and has for 20 years based on Moores law and that exponential graph. The world will change faster every year. We will not be able to keep up. We have probably passed the point where this is reversible. I think we are doomed as a species, but we might just cure all diseases and some death. It might be utopia for a while. Guess we get the privilege of seeing this play out, however it goes.

William O'Leary

James, do we need to covert RNDR (ERC20) to Render? Can you please talk about this. I'm confused

Suzi Gemini
>>143515750

The problem is they don’t understand use it or lose it…. Our bodies atrophy. Bones reabsorb etc. if neurons can’t meet up with their buddies to make new synapses they die.

Apoka Ta Stasis

I have put off writing for years, always thinking it was something I would do when I retired. The introduction of GPT actually lit a fire to my keyboard and I’ve began writing my stories. I want to get my ideas on paper before these language models take over the art space. My hope is that most people will reject consuming AI produced material such as books, music, and movies … but my fear is that they won’t. So anyway, I am going to hammer out as many of my ideas as I can before the bots take over art.

BerryDingle4U
>>143521514

One render is on Solana chain and one is on etherium (ERC-20). I only hold the render on the Solana chain

Mitchell
>>143530090

Love it! The demand for genuine human art creation can only increase, right?

TheWorldEconomicForum

“Once men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to enslave them.”- Frank Herbert

Aegis87
>>143515802

The jump from GPT-3 -> GPT-4o, and the impending release of GPT-5 has convinced me that there is a very high likelihood of seeing level 4 in that time frame. I believe that both Yann Lacunn, and Andrew Ng have recently commented that collaborative models (multiple agents) working together can give you substantial improvements over all existing single models. As we see hardware, and algorithmic advancements in the next couple years its likely we'll be able to run 100s - 10,000s of agents working together, which may very well approximate level 4 behavior in short order.

Uba_nooba
>>143511804

People are not positioned correctly. Technology is already self replicating and replacing peoples jobs

subconscious

https://youtu.be/AAqnCYV5hkg?si=LaJkJsT4pFVFflZK implications of this is probably better than any other current technology potential. The richest and smartest man in the world whose successfully made humanoid robots (Optimus), that have access to unlimited energy (solar) & energy storage (megapack), making the best CPU chips in the world for their own tech (dojo) for the greatest processing, with the fastest & best connectivity Internet (starlink), with neurachips that can be precisely implanted into humans ultimately destined for healthy humans as performance enhancements to create cyborgs (neuralink). Which means a cyborg sitting anywhere can control a whole robot army which has access to interplanetary rockets (SpaceX). And he has access to the best real time info with X (most popular social media now). So his robots will have real time data. And he's willing to interfere in wars - eg. providing starlink access to Ukraine for their drone attack network He's the next Tony Stark Most of all this is CURRENTLY possible and getting better at a 10x magnitude with every update (happening on less than a yearly basis). And these companies all have the smartest people in the world trying to get in. Its more competitive than Harvard, Stanford, NASA, or Google.