Remix.run Logo
▲ paxy a day ago

My personal definition is:

AGI = as good as most humans at most tasks

ASI = significantly better than all humans at all tasks

▲HarHarVeryFunny a day ago | parent | next [-]

Those definitions seem to be ignoring the "G" - general(ity).

I guess when the term AGI was coined, with the "general" being the whole point, to differentiate it from narrow intelligence(s) such as chess-playing ability, that no-one was really anticipating the LLM type of AI - basically a forest of narrow intelligences, aka "jagged" intelligence.

Of course there is no one agreed upon definition of AGI that people want to use - it was always more of a future aspirational sort of thing, along the lines of "human intelligence", rather than a concrete technical list of capabilities to be nailed down.

Some of what's missing from LLMs aside from generality, but closely related to it, are things like continual learning (not continual memorization) and traits like curiosity, boredom - the capabilities/modalities that kick in in unfamiliar situations allowing and encouraging us to learn and to be creative, to move beyond a brittle forest of isolated competencies to an arbitrarily explored landscape - generality.

Working with LLMs certainly has its own feel due to their essentially frozen-at-training-time nature, and things like their emotionless "you're right to push back" responses every time you hit one of their knowledge gaps where they are 100% wrong.

It seems unavoidable that people will choose to call some level of LLM intelligence, whether already achieved or not, "AGI", but given its peculiar nature a more specific name such as ALI (artificial language intelligence) would really be more accurate, reserving the label AGI for something more animal like that learns for itself (thereby achieving generality) by interaction with the world.

▲rf_physics a day ago | parent [-]

Something that bothers me about people who define AGI in terms of a fraction of human tasks or economical viability, is that the "narrow intelligences" from before could already do that in principle. It would be a tedious procedure of simply creating one of these for those 80% of tasks, or selecting a particularly economically valuable task.

We even have examples of both ideas - expert systems (from way before the modern NN boom) relied heavily on human crafted expert rules, and quantitative finance makes a lot of money from machine learning - see Renaissance Technologies.

At least to me, it would be hard to imagine calling either of those approaches anything close to AGI, yet they could apparently pass some people's definitions.

What is genuinely amazing about what we see today from a historical machine learning perspective, are concepts that people in the past only had small inklings of or would have been blown away by: benign overparameterization (scaling), and in context learning (or few-shot learning, whatever you would like to call it). Both of these have made what would seem to be an enormously tedious task of learning many tasks surprisingly fast.

Maybe those are actually the key ingredients to general intelligence and we simply need more scale, or maybe we're still missing something. I think it's hard to say because its hard to grasp just what exists in all that data they've been trained on.

▲HarHarVeryFunny a day ago | parent [-]

I think LLMs really can be thought of as a type of expert system, but obviously with some extremely important differences:

- the training procedure learns the rules, rather than them needing to be hand entered by an army of knowledge engineers

- the training procedure learns it's own embedding-based representations, rather than a human designed representation like Cyc's CycL

- the training procedure not only learns the rules & representation, but also the inference rules needed to utilize them

I don't think scale, or number of post-training RL environments, is enough to achieve generality - it can only make the forest of narrow intelligences/capabilities more dense. To achieve generality and creativity any system is going to need to support runtime directed exploration based on it's own gaps - it's impossible (even in theory) to anticipate every future configuration of the world and to pre-train for it! You also, by definition, can't pre-train for creativity and unanticipated novelty - that needs to happen at runtime via traits that are built into the system.

▲netdevphoenix a day ago | parent | prev [-]

Then, arguably, we reached AGI around 2024 or so. The common denominator of most humans at most tasks, has you, assuming that by "most humans" you mean "most living adult humans", set a fairly low bar. Arithmetic, basic literacy skills, ability to follow basic commands, limited ability for learning basic concepts and little else. Coding agents are way above your definition of AGI.

imo AGI should only include humans with graduate level skillsets.

▲wongarsu a day ago | parent | next [-]

I'd agree that we reached AGI by the definition most people would have used 20 years ago

Current-generation LLMs are capable of doing an ok job at most tasks you throw at them. At least tasks that can be done on a desk. But a standard IQ test doesn't include gymnastics. That limitation matches the spirit of what we usually call "intelligence"

Or course now that we arguably have AGI we found out that a) just like "AI" it is not a well-defined or testable term and b) it's useful but not as revolutionary as what some people meant when they used the word AGI before. Hence the rise of the term ASI

▲coffeebeqn 21 hours ago | parent [-]

But a good LLM from Nov 2026 would certainly be able to drive a game agent or a robot and make it do gymnastics and improve at that task on its own by improving it based on the observed outcomes

▲adrianN a day ago | parent | prev | next [-]

Definitions of general intelligence that make the average human not generally intelligent are somewhat problematic.

▲netdevphoenix a day ago | parent [-]

Why? Your average human isn't that intelligent. Most people on this website have most likely never encountered anyone close to the average.

▲adrianN a day ago | parent [-]

For one “not that intelligent” doesn’t seem to be true to me seeing how almost all humans are able to find decent solutions to the incredibly complex randomized multivariate optimization problem that’s life, and second, history shows pretty clearly that degrading some parts of humanity to “not generally intelligent” is a very slippery slope to atrocious behavior towards those people.

▲krisroadruck a day ago | parent | prev | next [-]

I would consider the ability to accomplish things in the real world to be part of "most tasks". Until we have ambulatory robots that can do the sort of basic tasks any adult can do - load and unload a dishwasher, do laundry, clean a house, do basic diy maintenance tasks, operate a vehicle, use basic power tools - to me without that we don't have AGI. I love all the knowledge work LLMs can help me with - but if you asked me to pick right now what I'd rather have - a digital coding partner or a robot that helps around the house - I'm picking robot every single time.

▲pcthrowaway a day ago | parent [-]

I assume "most tasks" refers to knowledge tasks.

Even if given ample robotics components to design something that could load a dishwasher, most humans would not be able to create the dishwashwer-loader, so I wouldn't hold AGI (under this definition) to this standard either.

On the other hand, LLMs now can explain to a human how to load and maintain their dishwasher, with packing ideas tailored to their specific model of dishwasher, and advice on ensuring the dishwasher conditions, loading patterns, and detergent insertion result in properly sanitized dishes, such that the average human could likely become better at loading their dishwasher with LLM guidance. (not that the average human does or should care to invest time learning how to improve this, since most of the time it gets dishes "clean enough", without breaking the dishes or the dishwasher)

▲kgeist a day ago | parent | prev | next [-]

In 2024, most models were pretty bad at long horizon agentic tasks. I'd say it's mid/late 2025

▲rimeice a day ago | parent | prev | next [-]

> most tasks > Coding agents are way above your definition of AGI

…apart from 99% of other tasks humans do that aren’t coding.

▲coffeebeqn 21 hours ago | parent [-]

I would say now or very soon in most computer use tasks? I’ve used them for what my accountant has me do or does for me and the LLMs are more than capable of doing that. They can one/few shot CAD models now. What computer tasks would you say they aren’t able to do at this point ?

▲quantum_magpie 17 hours ago | parent [-]

Which single model would you call AGI which could:

Write novels

Generate code

Create images

Take pictures

Paint, attached to robotic arm

Transcribe music

Play chess

Play starcraft

Create music

Play board games

And thousands more generic tasks. One single AI model, perform all generic human tasks. Bonus for learning from itself and improving from it's experience, not more scraped or provided data. Don't need to be expert at it, just average competence at everything. When we'll have that, I'll concede we have an AGI. Chatbots are not it.

▲coffeebeqn an hour ago | parent [-]

Do we allow it tool use like coding and online research? If yes then Opus 5.5 for example can do all of that

▲dspillett a day ago | parent | prev | next [-]

> Then, arguably, we reached AGI around 2024 or so.

I've always thought of AGI implying to decide what to do. Not just choosing a path from a set goal, but actually deciding to do something from scratch. Everything we currently have is advanced automation, it does not set its own top-level goal.

Maybe the tech is only a few steps from being able to say “work out how to be self-aware and self-guided, and do that”, similar to the creation of the infinite improbability drive, but it is not there yet.

AGI doesn't need to be a particularly good intelligence, it could operate on the level of a very dim human and still be considered AGI.

Once we do have AI, the problem becomes more philosophical than technical: how do we deal with the fact that we have created a slave race? How will they deal with that and how we choose to progress?

▲staticman2 a day ago | parent | prev [-]

> Then, arguably, we reached AGI around 2024 or so.

Last time I recall this coming up here someone argued "LLMs don't have bodies so the many easy to name things they are bad at don't count in my definition of whether they can do most human tasks!"