Great Talk. Biggest challenge is to find a cheap chemical material acting as an accumulator to solve our renewable energy storage problems. An energy storage device from plastic waste found by AI would be the greatest achievement.
The greatest problem facing humankind today is the large volume of misinformation imbedded in our culture driven ideologies. Our best hope is that AI research will discover a rational system for testing alternative ideas on how to educate ourselves to increase our collective knowledge.
Observation: I find I must always proof read material which has word (or even context) completion to see if it coveys the meaning I intended. It seems to be even more true with LLM. Humans are still required to provide the intelligence factor.
Modeling the future of education after TikTok seems like an absurdly idiotic proposition. The war for attention cannot be won (except for corporations) by submitting to the hyperstimuli paradigm. Everyone needs to learn how to direct their attention towards seemingly boring subjects. I think the potential of AI in education as it stands today, lies in providing easier access to personalized help, what if the teachers were to record the questions posed by students and making a data set of the answers, making personal help avaliable outside the classroom context. However im not sure as to how not to diminish the individual problem solving skills of children in this way, maybe a model that only gives hints/directions toward discovering a certain solution could be built?
This episode was really engaging. Loved it.
Great Talk. Biggest challenge is to find a cheap chemical material acting as an accumulator to solve our renewable energy storage problems. An energy storage device from plastic waste found by AI would be the greatest achievement.
Q. How do we extract knowledge from the large volume of information consolidated by large language models?
Fun talk! I do find PyTorch more user friendly than TF. Also the ease of installation of PyTorch over TF 😬 is something I’ve heard from people
Try to deploy it in containers in the cloud, and you will change your mind
The greatest problem facing humankind today is the large volume of misinformation imbedded in our culture driven ideologies. Our best hope is that AI research will discover a rational system for testing alternative ideas on how to educate ourselves to increase our collective knowledge.
Observation: I find I must always proof read material which has word (or even context) completion to see if it coveys the meaning I intended. It seems to be even more true with LLM. Humans are still required to provide the intelligence factor.
AI benefits: voice recognition, recommender systems, vehicle driving assistance, information moderation,
Modeling the future of education after TikTok seems like an absurdly idiotic proposition. The war for attention cannot be won (except for corporations) by submitting to the hyperstimuli paradigm. Everyone needs to learn how to direct their attention towards seemingly boring subjects. I think the potential of AI in education as it stands today, lies in providing easier access to personalized help, what if the teachers were to record the questions posed by students and making a data set of the answers, making personal help avaliable outside the classroom context. However im not sure as to how not to diminish the individual problem solving skills of children in this way, maybe a model that only gives hints/directions toward discovering a certain solution could be built?