r3alist
Elite Member
GoI's AI strategy is unique and vastly different from the current players.
For the application layer (top layer), they are already preparing the large tech companies, professionals, students and researchers to take a prompt pivot.
For the second layer (the models), they are not looking to develop large and hugely costly LLMs that can handle almost everything, rather looking to develope 50-100 smaller and specific model clusters that can cumulatively handle 90-95% of the requirement making it extremely affordable.
For the third layer (chips), it's still dependent on others, though trying to catch up within 5 to 7 years.
For the fourth layer(data centres, GPUs, cloud and other infra), they are doing a public private partnership under govt umbrella to develop and make the infrastructure available to all at low and in some cases no cost.
For the fifth layer(energy), much work still remains.
I wish you would share more details, so the application layer they are just preparing, pivoting to what?
For the second layer, what use cases are you actually developing for, if you are just saying a smaller model with fewer parameters, well this is what deep seek already done.
For the third layer, so you are dependent on chips from external parties.


