The Daily Index — Wednesday, November 8, 2023 · TLDR Index

Across today’s newsletters

Today’s Index

TLDR

THE OPENAI KEYNOTE

OpenAI recently held its first-ever developer conference. This article presents an analysis of the keynote and its main announcements. The keynote was well-presented and entertaining. The live demos went off without a hitch and leveraged the fact that they were live. All of the new products and features announced were immediately available rather than being available in the coming weeks or months, as is increasingly the case for events like I/O or WWDC.

stratechery.com · 24 min

TLDR AI

NVIDIA IS PILOTING A GENERATIVE AI FOR ITS ENGINEERS

Nvidia revealed the testing of a large language model named ChipNeMo to enhance chip designers' productivity during a keynote at the IEEE/ACM International Conference on Computer-Aided Design. Although not yet fully proven, ChipNeMo aids in scripting for design tools, summarizing bug reports, and serving as an experienced chatbot for designers.

spectrum.ieee.org · 3 min

TLDR Web Dev

WRITING A COMPILER IS SURPRISINGLY EASY (PART 1)

This article discusses the process of writing a code generator for a systems programming language with a focus on translating a high-level language into x86-64 assembly code. It explains how to represent integer literals, variables, negation, and addition in the compiler and provides code examples for each.

sebmestre.blogspot.com · 9 min

TLDR InfoSec

37 VULNERABILITIES PATCHES IN ANDROID SECURITY UPDATES

The first set of patches addresses 15 vulnerabilities in Android’s Frameworks and Systems. One critical information disclosure vulnerability can lead to local information disclosure with no additional execution privileges needed while the rest of the vulnerabilities are rated as high. The second set of patches addresses vulnerabilities in Arm, MediaTek, and Qualcomm components.

securityweek.com · 2 min

TLDR Crypto

USDC ISSUER CIRCLE MULLING IPO IN EARLY 2024

Circle, the second largest stablecoin issuer, is considering an IPO in early 2024 amidst improving market conditions in the space. This comes on the heels of crypto exchange Coinbase acquiring a stake in Circle. The plans also come during debates on the need for stablecoin regulation in the US. Currently, USDC has a market cap of $24.43 billion, while Tether, the largest stablecoin issuer, has a market cap of $86 billion.

cointelegraph.com · 2 min

TLDR

REDDIT IS STARTING TO TEST ITS OWN ADD-ONS FOR THE PLATFORM

Reddit is starting to test add-ons built through its Developer Platform. Its first add-ons include tools for mods to more easily monitor and remove comment threads, ban spammers and remove their content, and more. They will only be available to a small number of subreddit mods. The Developer Platform is still in beta - it is unknown when it will be generally available.

theverge.com · 2 min

TLDR

ELON MUSK’S NEURALINK IS WORKING ON A CHIP THAT WILL SOLVE BLINDNESS

Neuralink is planning to release a vision chip that will solve eye conditions that cause blindness within the next few years. The company has received approval to start recruiting humans for its first clinical trials. Its v0.9 chip aims to fix various ailments, including quadriplegia due to cervical spinal cord injury or amyotrophic lateral sclerosis (ALS). Neuralink is still waiting for regulatory approval for its first human tests.

teslarati.com · 2 min

TLDR

CRUISE CONFIRMS ROBOTAXIS RELY ON HUMAN ASSISTANCE EVERY FOUR TO FIVE MILES

Cruise uses a remote assistance team to help its autonomous fleet in certain situations. The company claims that remote assistance is being used about 2-4% of the time on average in complex urban environments. It says that the number is low enough that there isn't a huge cost benefit to optimizing much further. The remote assistants mostly provide wayfinding intel to the vehicle instead of controlling it remotely and more than 98% of sessions are answered within 3 seconds. It is an industry standard to have humans on call.

cnbc.com · 4 min

TLDR

ARE LANGUAGE MODELS GOOD AT MAKING PREDICTIONS?

This post presents the findings of a study that looks at whether GPT-4 can make good predictions. GPT-4's predictions are not well calibrated - its predicted probabilities are often over-confident. Its prediction ability varies depending on the category - for unclear reasons, GPT-4 is well-calibrated for sports questions but horrendously calibrated for personal questions. The model is better at making predictions for politics than for science or technology.

lesswrong.com · 5 min