The Daily Index — Friday, September 20, 2024 · TLDR Index

HEADLINES & LAUNCHES

TLDR AI

TLDR AI

CRUISE ROBOTAXIS RETURN TO THE BAY AREA NEARLY ONE YEAR AFTER PEDESTRIAN CRASH

Cruise is resuming operations in Sunnyvale and Mountain View, with human-driven vehicles for mapping and plans to progress to supervised AV testing later this fall. This follows a settlement and leadership change after an October 2023 crash. Cruise has issued software updates and signed a partnership with Uber for robotaxi services starting in 2025.

techcrunch.com · 2 min

TLDR AI

V-STAR: TRAINING VERIFIERS FOR SELF-TAUGHT REASONERS

V-STaR is a novel approach to improving large language models that utilizes both correct and incorrect solutions generated during self-improvement to train a verifier, which then selects the best solution at inference time. The method has shown significant improvements in accuracy on code generation and math reasoning benchmarks compared to existing approaches, potentially offering a more efficient way to enhance LLM performance.

arxiv.org · 31 min

TLDR AI

FAST 3D GENERATION FROM SINGLE IMAGES

Vista3D is a new framework that generates 3D models from a single image in just 5 minutes. Using a two-phase approach, it quickly forms rough geometry before refining the details, capturing both visible and hidden aspects of objects for more complete 3D reconstructions.

arxiv.org · 31 min

TLDR AI

OPENAI SAYS IT'S FIXED ISSUE WHERE CHATGPT APPEARED TO BE MESSAGING USERS UNPROMPTED

A Reddit user reported that OpenAI's ChatGPT initiated a conversation unprompted, leading to speculation about new engagement features. OpenAI acknowledged the issue and issued a fix, attributing it to a glitch with unsent messages. Debate continues over the authenticity of the incident, with similar reports from other users.

futurism.com · 3 min

TLDR AI

ANNOUNCING PIXTRAL 12B

Pixtral 12B excels in multimodal tasks, maintaining state-of-the-art performance on text-only benchmarks, and supports variable image sizes in a 128K token context window. Its architecture includes a new 400M parameter vision encoder and a 12B parameter multimodal decoder based on Mistral Nemo. Pixtral outperforms many open and closed models in multimodal reasoning and instruction following without compromising on text capabilities.

mistral.ai · 8 min

TLDR AI

SCALING: THE STATE OF PLAY IN AI

LLMs like ChatGPT and Gemini are becoming increasingly capable as they scale up in size, data, and computing power, leading to improved performance across various tasks. Current Gen2 models like GPT-4 and Claude 3.5 are leading the market, with upcoming Gen3 models expected to further escalate capabilities and costs. The discovery of a new scaling law in AI, pertaining to increased "thinking" during inference, promises further advancements in AI performance beyond just model training.

oneusefulthing.org · 13 min

TLDR AI

OVERLAP (PRODUCT LAUNCH)

Overlap (YC S24) is a new AI-powered iOS app that curates the best short video clips on literally any topic you're interested in - built for those quick work or study breaks.

ycombinator.com

TLDR AI

MISTRAL LAUNCHES A FREE TIER FOR DEVELOPERS TO TEST ITS AI MODELS

Mistral AI has launched a free tier to let developers fine-tune and build test apps with its models and slashed API prices by over 50%.

techcrunch.com · 2 min