Sretto

2/29/24 2:27PM

Here the three trends I felt for this year(2024):

Continued study for protecting the planet Earth

Last year, experiment was conducted to deflect astroid away from Earth. I think we can see more studies being conducted to protect Earth from potential events in the space this year and beyond. The current popular theory about the universe is that it's been expanding for billions of years. If the universe expands indefinitely like letter "/" (forward slash), could there be any adverse consequences affecting Earth? Or could there be any chances that the universe is going like letter "A", first expanding then reducing? Or could the universe be going like letter "M", in waves of expansions and reductions? In any case, more studies on the potential change of the speed of the big bang, and its effect on Earth should be considered to be included in further study of Earth protection, I think.

Open source hardware for maching learning

Last year, many hardware architectures for training LLM and conducting inferences emerged. This year, we might see some interesting open source hardware architectures for training and inferencing. The open source hardware for machine learning could go really well with open source CPU architectures like RISC V in the same packaging.

Large language models would accelerate the adoption of specialized robots in our daily life

Today, there're specialized robots in a number of fields like manufacturing and space exploration. Large language models could make robots more predictable and easier to command for ordinary people. I feel more specialized robots could show up this year in transportation, food service, and health care to help ordinary people to go on with the daily life and to provide temporarily disabled people one additional way to undergo individualized recovery therapy, due to the fact that specialized robots can be easier to instruct via LLM.

Happy new year!

Feng Zhou



About Us|Corporate Responsibility|Contact|Advertising Program

Sretto Home

© 2026 Sretto