TLDR AI 2024-07-02

deletion requests, Incogni will do it for you [4]. Use this link

to get 60% off by using code TLDRAI.

❌ Delete your personal data today

MICROSOFT'S MUSTAFA SULEYMAN SAYS HE LOVES SAM ALTMAN, BELIEVES HE'S SINCERE ABOUT AI SAFETY

Mustafa Suleyman, CEO of Microsoft AI, praised OpenAI's CEO Sam Altman and discussed the collaborative yet independent relationship between Microsoft and OpenAI in a recent interview at the Aspen Ideas Festival. Suleyman emphasized the need for a dialogue around AI safety and regulation rather than perceiving the development of AI as a race. He also expressed interest in AI cooperation with China and the potential role of AI as a teaching aid in classrooms.

WHEN THE TERMS OF SERVICE CHANGE TO MAKE WAY FOR A.I. TRAINING

Tech companies like Google and Meta are updating their privacy policies to use public and possibly private user data to train AI systems as they navigate a complex landscape of privacy laws and user consent. Backlashes have occurred as users and creators are fearful about the use of their content in training AI that could potentially replace them. The tensions highlight emerging challenges in AI development, data privacy rights, and the balance between innovation and ethics in the tech industry.

MEET FIGMA AI

Figma AI is a new suite of AI-powered tools for Figma that helps designers with tasks such as visual search, asset search, text editing, image editing, prototyping, layer renaming, and design generation. These features are powered by third-party AI models and are free to use during the beta period.

SYNTHETIC DATA CREATION WITH 1 BILLION PERSONAS

This work treats web text as coming from a persona and then can condition on that persona to dramatically improve downstream task performance. The researchers find a jump of 20% points on MATH.

DEEPDIVE: SCALING UP WITH MUP

A really great and surprisingly accessible post about tuning hyperparameters as you scale up model and dataset size.

WEBPAGE CODE GENERATION

Researchers have developed a new process to improve Web2Code instruction tuning. It involves creating new webpage image-code pairs, refining existing webpage code generation data, creating new text question-answer pairs, and refining existing webpage understanding data.

panelists!

BLOCK TRANSFORMER FOR IMPROVED INFERENCE

This repository introduces a new type of Transformer with a dramatically reduced KV cache size. It hasn't been proven at scale, but it can - in theory - match regular Transformer performance.

COMPOSIO

Composio equips agents with well-crafted tools, empowering them to tackle complex tasks.

UNSUPERVISED SAM FOR AUTOMATIC IMAGE SEGMENTATION

Unsupervised SAM (UnSAM) is a new model for whole-image segmentation that eliminates the need for human annotations. Using a divide-and-conquer approach, UnSAM identifies hierarchical structures in visual scenes by combining top-down and bottom-up clustering methods.

AI SCALING MYTHS

The future potential of LLMs reaching AGI through scaling alone is unlikely. While scaling has shown improvements in model capabilities, it primarily enhances perplexity rather than emergent abilities. Acquiring high-quality training data is becoming increasingly challenging.

ANOMALY DETECTION WITH SCENE-SPECIFIC FOCUS

Researchers have introduced a new anomaly detection problem that targets 'odd-looking' objects relative to others in a scene. Unlike traditional methods, anomalies here are scene-specific and identified through multiple viewpoints.

A DISCUSSION OF DISCUSSIONS ON AI BIAS

Discussion on the nature of AI bias has intensified, with critics arguing that the way models like Playground AI sometimes alter a user's ethnicity in images demonstrates biases in ML. However, some users deny this is a bug or relevant bias, citing examples where Asian features are overrepresented. The debate extends to broader implications of AI bias across industries. It is a complex issue with no simple resolution.

FOLLOWING LENGTH CONSTRAINTS IN INSTRUCTIONS

Most models don't follow requests for length (e.g., fewer than 40 words). This work shows how you can tune them to do so.

THE SHAPE OF INFORMATION

This article explains how to maximize limited resources using binary logic.

SCREEN PIPE

Turn your screen into actions using LLMs.