✨ Edition 14 : Last 15 Days in AI ✨

See GPT results in Google Search.

😲 Chatgpt result in Google Search

Lets look the source:

We are about to live in tech feudalism - That’s the thought of the day. Period!!!

🤞 New Light: Language Modeling Is Compression

It has long been established that predictive models can be transformed into lossless compressors and vice versa. Incidentally, in recent years, the machine learning community has focused on training increasingly large and powerful self-supervised (language) models. Since these large language models exhibit impressive predictive capabilities, they are well-positioned to be strong compressors. In this work, we advocate for viewing the prediction problem through the lens of compression and evaluate the compression capabilities of large (foundation) models. We show that large language models are powerful general-purpose predictors and that the compression viewpoint provides novel insights into scaling laws, tokenization, and in-context learning. For example, Chinchilla 70B, while trained primarily on text, compresses ImageNet patches to 43.4% and LibriSpeech samples to 16.4% of their raw size, beating domain-specific compressors like PNG (58.5%) or FLAC (30.3%), respectively. Finally, we show that the prediction-compression equivalence allows us to use any compressor (like gzip) to build a conditional generative model.

In their study, the researchers evaluated the compression capabilities of LLMs using vanilla transformers and Chinchilla models on text, image, and audio data. As expected, LLMs excelled in text compression. For example, the 70-billion parameter Chinchilla model impressively compressed data to 8.3% of its original size, significantly outperforming gzip and LZMA2, which managed 32.3% and 23% respectively.

However, the more intriguing finding was that despite being primarily trained on text, these models achieved remarkable compression rates on image and audio data, surpassing domain-specific compression algorithms such as PNG and FLAC by a substantial margin. 

🔜 ChatGPT announces voice and image capabilities

  • Voice only on mobile: ChatGPT will be able to speak and understand speech.

  • Image input on all platforms: you will be able to upload images.

  • Over the next 2 weeks: roll out to Plus and Enterprise users.

💰 NVIDIA, The Most Recent Influential VC In Generative AI

Last week, AI news was dominated by Databricks' massive $500 million round. Who was at the center of that fundraising? NVIDIA. This news is only surprising if you haven't been paying attention to NVIDIA's recent investment spree. The recent additions to the portfolio feature some of the hottest companies in the AI space. Take a look and judge for yourself:

Judge yourself!!

Details is from The Sequence.

👃 Neural nets can smell now, too!!

Mapping molecular structure to odor perception is a key challenge in olfaction. Here, we use graph neural networks (GNN) to generate a Principal Odor Map (POM) that preserves perceptual relationships and enables odor quality prediction for novel odorants. The model is as reliable as a human in describing odor quality: on a prospective validation set of 400 novel odorants, the model-generated odor profile more closely matched the trained panel mean (n=15) than did the median panelist. Applying simple, interpretable, theoretically-rooted transformations, the POM outperformed chemoinformatic models on several other odor prediction tasks, indicating that the POM successfully encoded a generalized map of structure-odor relationships. This approach broadly enables odor prediction and paves the way toward digitizing odors.

One-Sentence Summary 

An odor map achieves human-level odor description performance and generalizes to diverse odor-prediction tasks.

Competing Interest Statement

The authors have declared no competing interest.

**

I will publish the next Edition on Sunday.

This is the 14th Edition, If you have any feedback please don’t hesitate to share it with me, And if you love my work, do share it with your colleagues.

It takes time to research and document it - Please be a paid subscriber and support my work.

Cheers!!

Raahul

**

Reply

or to participate.