@pj

:-) When I pointed out the analogy between #LLM training and biological #evolution, I wasn’t referring to biology’s ability to evolve language, but its capacity to develop almost anything at all: given enough tuneable parameters and a system for inheritance with variation under selective pressure, we get molecular rotors, nanofabrication, diffraction gratings for colours, magnetic compass, eyes, ears, labyrinths, social collaboration … and even language.

The number of tuneable parameters in the human genome is almost two orders of magnitude smaller than the number of parameters in GPT-3 (and less precise: only 2bit / four nucleotides). And training a language model means to follow a trajectory in a high-dimensional parameter space, just like evolution is a trajectory in genetic-sequence-space. (The technical difference is that the former is a directed walk along a gradient, while the latter is a random walk under selection).

And what happened is : when trained on token-prediction, LLMs started showing emergent aspects of “intelligence”.

Quanta just ran an article on that five days ago: https://www.quantamagazine.org/the-unpredictable-abilities-emerging-from-large-ai-models-20230316/

… and Google’s Jason Wei, who has a number of articles on arXiv on emergence, lists emergent abilities on his blog: https://www.jasonwei.net/blog/emergence

Molecular biology is one of my areas of expertise, I am not surprised this is happening: how could it not? But then again, to actually see such emergence is profound.

#Sentientsyllabus #ChatGPT #GPT4 #Bard #ai

The confluence of this technology with the information ecosystem that we described in our paper from a couple of years could be an epistemic catastrophe.

https://www.pnas.org/doi/10.1073/pnas.2025764118

I’m coming to think that releasing these tools was a reckless act with the potential to generate negative externalities we have barely started to imagine.

The threat isn’t rogue superintelligence. It’s bullshit at unprecedented scale, reflected back upon itself and iteratively amplified.