SingularityNet had a massive growth since ChatGPT came to fame and the project didn't suffer the last crypto crash making it somewhat uncorrelated to BTC and the rest of crypto market. However, in terms of technology framework like TensorFlow with models running on clusters of machine offers much better performance than models built on blockchains. It is unclear if this limitation will ever be overcome, so practical implementations of ML and AI technologies on top of SingularityNet may never become commercially viable. In theory ML models that benefit from being parallelized on GPUs could benefit from being parallelized on blockchain as well, but the technical difficulty of doing so increase exponentially. I'm not aware of any commercial applications running on SingularityNet so it is still an experiment/prototype after 7 years in the making.

There are in theory other advantages of building AI and ML on blockchain in regards to democratization of models and datasets, but is that enough? Democratization of data is a lot more viable and plausible to be achieved on blockchain than running the entire layer of intelligence on top.

Another disturbing aspect of SingularityNet is in regards to promotional material that indicate that building an AI on a blockchain somehow magically turn this AI into an Artificial General Intelligence (AGI) which is the holy grail attempted to reproduce human-level of intelligence and eventually giving birth to an intelligence that can be far superior than that of a human being with it's biological limitations. For one thing, despite the movie industry, we might still be very far from achieving AGI, and another layer of difficulty from achieving it on a decentralized infrastructure.

SingularityNET (AGIX) is a cryptocurrency and operates on the Ethereum platform

SingularityNET

SingularityNet had a massive growth since ChatGPT came to fame and the project didn't suffer the last crypto crash making it somewhat uncorrelated to BTC and the rest of crypto market. However, in terms of technology framework like TensorFlow with models running on clusters of machine offers much better performance than models built on blockchains. It is unclear if this limitation will ever be overcome, so practical implementations of ML and AI technologies on top of SingularityNet may never become commercially viable. In theory ML models that benefit from being parallelized on GPUs could benefit from being parallelized on blockchain as well, but the technical difficulty of doing so increase exponentially. I'm not aware of any commercial applications running on SingularityNet so it is still an experiment/prototype after 7 years in the making. 

There are in theory other advantages of building AI and ML on blockchain in regards to democratization of models and datasets, but is that enough? Democratization of data is a lot more viable and plausible to be achieved on blockchain than running the entire layer of intelligence on top.

Another disturbing aspect of SingularityNet is in regards to promotional material that indicate that building an AI on a blockchain somehow magically turn this AI into an Artificial General Intelligence (AGI) which is the holy grail attempted to reproduce human-level of intelligence and eventually giving birth to an intelligence that can be far superior than that of a human being with it's biological limitations. For one thing, despite the movie industry, we might still be very far from achieving AGI, and another layer of difficulty from achieving it on a decentralized infrastructure.

SingularityNET AGIX

Decentralised AI marketplace