News for hackers. Stuff that matters.
Re:Genie 3: A new frontier for world models by casenmgreen on Tuesday August 05, 2025 @12:27PM (#44800177)
I may be wrong, but this seems to make no sense.

A neural net can produce information outside of its original data set, but it is all and directly derived from that initial set. There are fundamental information constraints here. You cannot use a neural net to itself generate from its existing data set wholly new and original full quality training data for itself.

You can use a neural net to generate data, and you can train a net on that data, but you'll end up with something which is no good.

Re:Genie 3: A new frontier for world models by scarmig on Tuesday August 05, 2025 @03:42PM (#44803250)
Humans are dependent on their input data (through lifetime learning and, perhaps, information encoded in the brain from evolution), and yet they can produce out of distribution information. How?

There is an uncountably large number of models that perfectly replicate the data they're trained on; some generalize out of distribution much better. Something like dreaming might be a form of regularization: experimenting with simpler structures that perform equally well on training data but generalize better (e.g. by discovering simple algorithms that reproduce the data equally well as pure memorization but require simpler neural circuits than the memorizing circuits).

Once you have those better generalizing circuits, you can generate data that not only matches the input data in quality but potentially exceeds it, if the priors built into the learning algorithm match the real world.

Re:Genie 3: A new frontier for world models by stavros on Tuesday August 05, 2025 @04:11PM (#44803623)
Humans produce out-of-distribution data all the time, yet if you had a teacher making up facts and teaching them to your kids, you would probably complain.
Re:Genie 3: A new frontier for world models by scarmig on Tuesday August 05, 2025 @04:23PM (#44803767)
Humans also sometimes hallucinate and produce non-sequitors.
Re:Genie 3: A new frontier for world models by suddenlybananas on Tuesday August 05, 2025 @04:39PM (#44803952)
Maybe you do, but people don't "hallucinate". Lying or being mistaken is a very different thing.
Re:Genie 3: A new frontier for world models by delusional on Tuesday August 05, 2025 @04:07PM (#44803571)
Computers aren't humans.

We have truly reached peak hackernews here.

Re:Genie 3: A new frontier for world models by neom on Tuesday August 05, 2025 @12:46PM (#44800500)
I might be misunderstanding your comment so sorry if so. Robots have sensors and RL is a thing, they can collect real world data and then processing and consolidating real world experiences during downtime (or in real time), running simulations to prepare for scenarios, and updating models based on the day's collected data. The way I saw it that I thought was impressive was the robot understood the scene, but didn't know how the scene would respond to it's actions, so it gens videos of the possible scenarios, and then picks the best ones and models it's actuation based on it's "imagination".
Re:Genie 3: A new frontier for world models by Demplolo on Tuesday August 05, 2025 @03:27PM (#44803039)
I actually think you can.

The LLM has plenty of experts and approaches etc.

Give it tool access let it formulate it's own experiments etc.

The only question here is if it becomes a / the singularity because of this, gets stuck in some local minimum or achieves random perfection and random local minimum locations.

Re:Genie 3: A new frontier for world models by schmidtleonard on Tuesday August 05, 2025 @12:43PM (#44800448)
We are miles away from the fundamental constraint. We know that our current training methodologies are scandalously data inefficient compared to human/animal brains. Augmenting observations with dreams has long been theorized to be (part of) the answer.
Re:Genie 3: A new frontier for world models by vanviegen on Tuesday August 05, 2025 @05:46PM (#44804822)
> current training methodologies are scandalously data inefficient compared to human/animal brains

Are you sure? I've been ingesting boatloads of high definition multi-sensory real-time data for quite a few decades now, and I hardly remember any of it. Perhaps the average quality/diversity of LLM training data has been higher, but they sure remember a hell of a lot more of it than I ever could.

Re:Genie 3: A new frontier for world models by hnuser123456 on Tuesday August 05, 2025 @12:52PM (#44800582)
It's feasible you could have a personal neural net that fine-tunes itself overnight to make less inference mistakes in the future.
Re:Genie 3: A new frontier for world models by exe34 on Tuesday August 05, 2025 @02:26PM (#44802143)
Any idea how humans do it? Where do they get novel information from?
Re:Genie 3: A new frontier for world models by tim333 on Tuesday August 05, 2025 @05:25PM (#44804547)
Humans can learn from visualising situations and thinking through different scenarios. I don't see why AI / robots can't do similar. In fact I think quite a lot of training for things like Tesla self driving is done in simulation.
Re:Genie 3: A new frontier for world models by thecupisblue on Tuesday August 05, 2025 @05:39PM (#44804718)
This is definitely one of the potential issues that might happen to embodied agents/robots/bodies trained on the "world model". As we are training a model for the real world based on a model that simulates the real world, the glitches in the world simulator model will be incorporated into the training. There will be edge cases due to this layered "overtraining", where a robot/agent/body will expect Y to happen but X will happen, causing unpredictable behaviour.I assume that a generic world agent will be able to autocorrect, but this could also lead to dangerous issues.

I.e. if the simulation has enough videos of firefighters breaking glass where it seems to drop instantaneously and in the world sim it always breaks, a firefighter robot might get into a problem when confronted with unbreakable glass, as it expects it to break as always, leading to a loop of trying to shatter the glass instead of performing another action.

Re:Genie 3: A new frontier for world models by fc417fc802 on Wednesday August 06, 2025 @02:20AM (#44808366)
Approximately speaking, you have a world model and an agent model. You continue to train the world model using data collected by the robot day-to-day. The robot "dreams" by running the agent model against the world model instead of moving around in the real world. Dreaming for thousands of (simulated) hours is much more efficient than actually running the physical hardware for thousands of wall clock hours.
Re:Genie 3: A new frontier for world models by roenxi on Wednesday August 06, 2025 @06:14AM (#44810110)
It is possible - for example, getting a blob of physics data, fitting a curve then projecting the curve to theorise what would happen in new unseen situations. The information constraints don't limit the ability to generate new data in a specific domain from a small sample; indeed it might be possible to fully comprehend the domain if there is an underlying process it can infer. It is impossible to come up with wildly unrelated domains though.
Re:Genie 3: A new frontier for world models by kannanvijayan on Wednesday August 06, 2025 @10:09AM (#44812254)
The benefit of these AI-generated simulation models as a training mechanism is that it helps add robustness without requiring a large training set. The recombinations can generate wider areas of the space to explore and learn with but using a smaller basis space.

To pick an almost trivial example, let's say OCR digit recognition. You'll train on the original data-set, but also on information-preserving skews and other transforms of that data set to add robustness (stretched numbers, rotated numbers, etc.). The core operation here is taking a smallset in some space (original training data) and producing some bigset in that same space (generated training data).

For simple things like digit recognition, we can imagine a lot of transforms as simple algorithms, but one can consider more complex problems and realize that an ML model would be able to do a good job of learning how to generate bigset candidates from the smallset.

Re:Genie 3: A new frontier for world models by robotresearcher on Wednesday August 06, 2025 @01:19PM (#44814864)
AlphaGo would seem to be a conceptually simple counter example.