Re: AI KR , foundation models explained (talking about slippery things

Hey, Paola & Dave, your knowledge far exceeds my own on these complex topics.  However, with the aid of Claude.ai and ChatGPT, I now have a better understanding of how they relate to human values and objectives.
Claude.ai's conclusion:

... properly aligning the fast and slow cognitive capabilities of LLMs with human values is crucial for beneficial and robust AI systems that can achieve our objectives while respecting key ethical principles. A combination of training data curation, reinforcement techniques, human collaboration, value learning and scalable oversight provides a pathway to this goal.https://claude.ai/chat/875f38e4-24fb-4a30-8a97-9d95484fb63c


ChatGPT:

 Complex Problem Solving: For objectives that require deep thinking, analysis, and planning, System 2-like processing in LLMs is essential.
Strategic Planning: LLMs can assist in strategic decision-making by analyzing large datasets, identifying trends, and providing comprehensive insights.https://chatgpt.com/share/de25536f-5f8d-4c07-b133-92ec81707e1f

Both Claude.ai and ChatGPT are becoming fluent in explaining why StratML is highly relevant to the latter point.  Failing to capitalize on such potentials would seem to require extrodinary degrees of inattentional blindness and artificial ignorance, in support of maintenance of the prevailing politically motivated powers-that-be.
See also:  AI for Politics-Free Life?
Owen Amburhttps://www.linkedin.com/in/owenambur/
 

    On Saturday, June 8, 2024 at 05:00:12 AM EDT, Dave Raggett <dsr@w3.org> wrote:   

 Training foundation models for LLMs is kind of like getting them to learn about everything all at once, all mixed up. This  works thanks to the magic of gradient descent and back propagation, and addresses the challenge that understanding every day sentences requires a good grasp of common sense knowledge, creating a chicken and egg problem.
Back propagation is very slow (look at typical values for the learning rate) and very different from how humans learn. We are able to learn from single examples, and get by on a very tiny fraction of the data that LLMs require for foundation models. Chomsky referred to this as the “poverty of the stimulus”.
During childhood, our schooling introduces knowledge in a carefully organised approach with new knowledge layered on top of previously learnt knowledge.  Our grasp of common sense comes from a blend of everyday experience and what we are schooled.
In the last ten years AI has come a long way, but we are still to figure out how to mimic the economies of human learning. I am searching for the means for neural networks to memorise and generalise from sequences using single-shot learning. This means stepping away from back propagation to consider other, more biologically plausible approaches.  One paper that caught my eye combines slow learning for learning to learn, and fast learning for single-shot learning. In essence, this trains the network to learn quickly for a limited set of tasks.
Tomorrow’s AI will be very different from today’s as we gradually master quick learning and deliberative (Type 2) reasoning. Moreover, it will use a fraction of the power consumed by today’s energy hungry GPUs/TPUs.  Sparse spiking neural networks implemented with neuromorphic hardware will mimic the efficiency of the brain.  This is also likely to trigger a move away from back propagation.
There is a lot to look forward to.
Cheers,    Dave


On 8 Jun 2024, at 06:03, Paola Di Maio <paola.dimaio@gmail.com> wrote:
Okay, folks, I have been a bit AWOL, got lost in the dense forest of understanding following the AI KR pathIn related discussions, what are foundation models?
If you ask Google (exercise)  the answer points to FM in ML, starting with Stanford in 2018 etc etc etchttps://hai.stanford.edu/news/what-foundation-model-explainer-non-expertsGreat resources to be found online, all pointing to ML and nobody actually showing you the FMis in a tangible form (I remember this happened a lot with SW)
Apparently
that FM are actually not an actual thing, they are not there at all, they are like dynamic neural network architecture (no wonder they have been slippery all along) which is built by ingesting 
data on the internet

Foundation models are massive neural network-based architectures designed to process and generate human-like text. They are pre-trained on a substantial corpus of text data from the internet, allowing them to learn the intricacies of language, grammar, context, and patterns.
They are made of layers, heads and parameters

Coming from systems engineering, you know, with a bit of an existential background, I am making the casethat foundational models without ontological basis are actually the cause of much risk in AI
In case you people were wondering what I am up to, and would like to contribute to this workPlease pitch in
Paola 


Dave Raggett <dsr@w3.org>