Plans for new types of problem to test General Intelligence of upcoming new models like o3 ?

[Greatz08]

Interested to know if there are any new plans to introduce new questions to test general intelligence because recently o3 model which is about to be released in near future will be easily able to clear this current level of questions easily and i personally highly doubt that AGI is achieved so i think their could be better ways to trick and test intelligence of these upcoming new models from open ai and other sources so that we can further push them to improve their models to actually attain AGI :-)

[mi3law]

Why do you doubt AGI is achieved?

We're working on tests and benchmarks at aolabs.ai starting from the simplest examples of learning shared by children and animals, get in touch!

[Greatz08]

@mi3law i doubt because even tho they(open ai) achieved great results with their new upcoming model o3, still they didnt declared by themselves that AGI is 100% achieved now and they also had a conversation with ARC AGI owner i guess where they also said that now they have to develop more problems for these upcoming models and test on that and see if they are still able clear those tests or not and if they cant then it means that open ai did trained them specifically to clear current ARC AGI Problems.

I do know that they mentioned in their video that they didnt specially trained their model to solve ARC AGI problem still i and many others cant/will ever trust any tech giant that easily and there are many reasons for it which i wont discuss here but yeah it is what it is.

So i thought to ask regarding the same like if you all are working on new complex problems which can trick models in solving problems with advanced thinking too . What i think personally is that if they are able to solve those new complex problems then we can for sure say that they are close to AGI and then we need some different extreme+new level coding,mathematics,physics etc problems and if it is able to solve those with 'consistency' with proper thinking and without hallucinations then i guess we can say it reached 85-90% AGI , then after that if it can improve itself and can do some actual discovery then we can say 100% achieved .
(Again this is what i personally think and your opinion could be different and i have no issues with that )