Best of 2023

The blog posts Hacker News loved most in 2023. 6,299 posts.

Any topic
6272

Why Software Is Eating the World

Software is eating the world. More than 10 years after the peak of the 1990s dot-com bubble, a dozen or so new Internet companies like Facebook and Twitter are sparking controversy in Silicon Valley, due to their rapidly growing private market valuations, and even the occasional successful IPO. With scars from the…

6274

AI Written, AI Read cartoon

One piece of slang that has long embodied the short attention span Internet age is TL;DR, short for “too long; didn’t read.” With the explosion of generative AI tools, we’re rapidly entering the age of TL;DW: “too long, didn’t write.” A January survey from Fishbowl found that 40% of nearly 12,000 workers have used…

6285

Aligning language models to follow instructions

We’ve trained language models that are much better at following user intentions than GPT-3 while also making them more truthful and less toxic, using techniques developed through our alignment research. These InstructGPT models, which are trained with humans in the loop, are now deployed as the default language models…

6286

CLIP: Connecting text and images

We’re introducing a neural network called CLIP which efficiently learns visual concepts from natural language supervision. CLIP can be applied to any visual classification benchmark by simply providing the names of the visual categories to be recognized, similar to the “zero-shot” capabilities of GPT-2 and GPT-3.

6289

F-Droid Security Issues

F-Droid is a popular alternative app repository for Android, especially known for its main repository dedicated to free and open-source software. F-Droid is often recommended among security and privacy enthusiasts, but how does it stack up against Play Store in practice? This write-up will attempt to emphasize major…

6298

The Scaling Hypothesis

On GPT-3: meta-learning, scaling, implications, and deep theory. The scaling hypothesis: neural nets absorb data & compute, generalizing and becoming more Bayesian as problems get harder, manifesting new abilities even at trivial-by-global-standards-scale. The deep learning revolution has begun as foretold.