| Newsletter | https://deepchasm.substack.com/ |
| Website | https://www.aqonemaki.com |
| https://www.linkedin.com/in/aqonemaki/ | |
| Bluesky | @aqonemaki.bsky.social |
Just read an awesome essay on the Scholar's Stage comparing @peterthiel 's and Vaclav Smil's differing theses on technological stagnation.
I'm wondering whether nuclear weapons could have been the turning point between technological optimism to risk aversion?
Nuclear fission is the first general purpose technology that has clear, obvious, and potentially immediate, existential risks associated with it.
Well worth reading the full essay: https://scholars-stage.org/about-the-author/
Which companies are the best in the world at Product Management?
I'm thinking Apple would be there. But what about 3M? Nintendo? Stanley Black and Decker? Others?
The #CPU within a phone charger is 563 times faster than the Apollo 11 Guidance Computer, can store 1.78x more instructions, has a little over twice the RAM.
Phone chargers to the moon!
https://forrestheller.com/Apollo-11-Computer-vs-USB-C-chargers.html
A fun game to test prompt injection vulnerabilities in LLMs. Can you get the AI to reveal the secret password?
Stable Diffusion revolutionised image creation from descriptive text. GPT-2, GPT-3(.5) and GPT-4 demonstrated astonishing performance across a variety of language tasks. ChatGPT introduced such language models to the general public. It is now clear that large language models (LLMs) are here to stay, and will bring about drastic change in the whole ecosystem of online text and images. In this paper we consider what the future might hold. What will happen to GPT-{n} once LLMs contribute much of the language found online? We find that use of model-generated content in training causes irreversible defects in the resulting models, where tails of the original content distribution disappear. We refer to this effect as Model Collapse and show that it can occur in Variational Autoencoders, Gaussian Mixture Models and LLMs. We build theoretical intuition behind the phenomenon and portray its ubiquity amongst all learned generative models. We demonstrate that it has to be taken seriously if we are to sustain the benefits of training from large-scale data scraped from the web. Indeed, the value of data collected about genuine human interactions with systems will be increasingly valuable in the presence of content generated by LLMs in data crawled from the Internet.
Innovation is hard! To innovate well you must accept that failure is a likely outcome and celebrate it when it happens but also learn from the mistakes that were made.
