541
How to scale LLMs better with an alternative to transformers
Monarch Mixer: Revisiting BERT, Without Attention or MLPs
Topics /
AI & Machine Learning
The most popular blog posts about ai & machine learning on Hacker News. 789 posts.
542
Using GPT-4 to measure the passage of time in fiction
Large language models are valuable research assistants, especially when they refuse to follow instructions.
543
ChatGPT plugins now support Postgres and Supabase
Supabase recently contributed to the OpenAI Retrieval Plugin repo with a Postgres and a Supabase implementation to help developers build ChatGPT plugins using pgvector.
544
Replika AI: Your Money or Your Wife
Isn't there somebody you forgot to ask?
545
Role-playing with AI will be a powerful tool for writers and educators
Or, how well can GPT-4 simulate an acid trip in 1963?
546
Natural language is an unnatural interface
How can we use familiar design patterns to constrain the unwieldy state space of large language models?
547
Double Descent in Human Learning
The uncanny resemblance between double descent and a 50 year old theory from psychology
548
Workers with less experience gain the most from generative AI
Less-experienced contact center agents become more productive with assistance from generative artificial intelligence, a new study finds.
549
550
Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields
Neural Radiance Field training can be accelerated through the use of grid-based representations in NeRF's learned mapping from spatial coordinates to colors and volumetric density. However, these grid-based approaches lack an explicit understanding of scale and therefore often introduce aliasing, usually in the form…
551
Stability AI releases DeepFloyd IF, a powerful text-to-image model
DeepFloyd IF is a state-of-the-art text-to-image model released on a non-commercial, research-permissible license that allows research labs to examine and experiment with advanced text-to-image generation approaches. In line with other Stability AI models, Stability AI intends to release a DeepFloyd
552
Is AI killing the stock photo industry? A data perspective
When Stock Performer started in 2010, there were a lot of complaints that the stock industry was dying. The rush of the agencies to implement subscription models, triggered mostly by Shutterstock’s success, led to a lot of uncertainty about photographers’ income. After all, customers were able to buy images for mere…
553
AI is just someone else's intelligence
It’s been a long time since I’ve worked in the field of ML (or what some call AI), and we’ve come a long way from simple text classification to what’s being casually called generative AI today. While the technology has made many advances, the foundational concepts of machine learning have remained analogous over time…
554
Artificial intelligence just lost a leader
Plus a position-search announcement from NSF Roger Schank just passed away. Roger was a top leader of AI. I overlapped with him for my time at Yale. In 1974, he became a professor of computer scien…
555
Debugging a Mixed Python and C Language Stack
Debugging is difficult. Debugging across multiple languages is especially challenging, and debugging across devices often requires a team with varying skill sets and expertise to reveal the underlying…
556
557
558
559
560
RAG to Riches
Today I will be regaling you with tales of marvelous cherry-picked new developments in AI, from Sam Altman's recent Uno Reverse incident to bear spraying job-hungry engineers.
561
Like diffusion but faster: The Paella model for fast image generation
The ability to generate realistic images without waiting would unlock applications from engineering to entertainment and beyond. New work takes a step...
562
Weak-to-Strong Generalization
We present a new research direction for superalignment, together with promising initial results: can we leverage the generalization properties of deep learning to control strong models with weak supervisors?
563
GPT-4 is not getting worse
GPT-4 may have declined in some areas, but it hasn’t systematically done so. My tests were shit, and never proved much of anything.
564
Margaret Mead, technocracy, and the origins of AI's ideological divide
The anthropologist helped popularize both techno-optimism and the concept of existential risk
565
Our humble attempt at “how much data do you need to fine-tune”
Efficacy of OpenAI Fine-tuning API, Cost/Latency Considerations, Experiencing Catastrophic Forgetting, and...Getting Roasted by a Fine-tuned GPT-3.5
566
The Most Cited AI Papers in 2022
Who Is publishing the most Impactful AI research right now? With the breakneck pace of innovation in AI, it is crucial to pick up some signal as soon as possible. No one has the time to read everything, but these 100 papers are sure to bend the road as to where our AI technology is going. The real test of impact of…
567
LLM tech comes to Wolfram Language
Wolfram's Version 13.3 includes a subsystem around LLMs plus new functionality in integrals, special functions, finite fields, computational geometry, live highlighting in visualizations, AR & VR deployment. See examples for those and dozens of other enhancements.
568
569
FlashAttention: Fast Transformer training with long sequences
Transformers have grown deeper and wider, but training them on long sequences remains difficult. The attention layer at their heart is the compute and memory bottleneck: doubling the sequence length would quadruple the runtime and memory requirements.
570