Daily AI Newsletter: AI Will Become Mathematicians’ ‘Co-Pilot’

In today’s email:

🏆 ARC Prize, a $1,000,000+ prize pool competition to beat and open-source a solution to the ARC-AGI eval.
🇧🇷 Brazil hires OpenAI to cut costs of court battles
🪫 AI is poised to drive 160% increase in data center power demand
🧰 8 new AI-powered tools and resources. Make sure to check the online version for the full list of tools.

AI Will Become Mathematicians’ ‘Co-Pilot’

Mathematics is undergoing a transformation with the advent of automated proof checkers and artificial intelligence (AI). Terence Tao, a renowned mathematician from UCLA, discusses how these technologies are revolutionizing the field by enabling larger-scale collaborations without the need for personal trust between mathematicians. Proofs can now be broken into smaller components, each verified by programs like the Lean compiler, allowing contributions from numerous individuals. This shift is fostering a more collaborative and efficient approach to solving complex mathematical problems.

The development of comprehensive mathematical libraries, such as Lean’s mathlib, has made formalization of proofs more practical and accessible. These libraries contain fundamental theorems, facilitating the verification of new fields of mathematics. AI's potential to assist in formalizing proofs could significantly reduce the time and effort required for mathematicians, transforming the way they work. Instead of manually typing proofs, future mathematicians might explain their ideas to an AI, which would formalize them and prepare them for publication.

While AI is not yet capable of independently proving major theorems, it is expected to become a valuable tool for mathematicians, helping to break down complex problems and suggest possible solutions. Tao envisions a future where AI aids in mass-producing proofs, similar to industrial production, allowing human mathematicians to focus on more innovative and higher-level mathematical concepts. This collaboration between human intuition and AI efficiency could lead to unprecedented advancements in mathematics, making the field more dynamic and interconnected.

Two Forbes 30U30 Founders Transforming Mental Wellness

Forbes 30 Under 30 winners founded Aura to solve the $100B problem - mental wellbeing.

Aura has quickly grown to 8 million users & 100k+ paying subscribers, and attracted investments from legendary Silicon Valley VCs & executives from Spotify, Facebook, and Apple.

The best part? You can invest now before the opportunity ends in a few days.

Elon Musk withdraws lawsuit against OpenAI and CEO Sam Altman

Elon Musk withdrew his lawsuit against OpenAI and its co-founders, Sam Altman and Greg Brockman, in California state court on Tuesday. The dismissal came just a day after Musk publicly criticized OpenAI's new partnership with Apple. The case, initially filed in February, alleged breach of contract and fiduciary duty. Experts had previously noted the lawsuit's shaky legal foundation due to the absence of a formal written agreement signed by all parties involved.

A hearing was scheduled for Wednesday in San Francisco to consider whether the case should be dismissed as requested by the defendants. Musk had claimed that the original mission of OpenAI to develop artificial general intelligence (AGI) for the benefit of humanity had shifted towards a for-profit model controlled by Microsoft, a major shareholder. The lawsuit served as a reminder of Musk's role in OpenAI's creation, although legal experts questioned the strength of his case.

Musk's withdrawal of the suit comes in the context of his own AI ventures. Last year, he launched xAI, an OpenAI competitor, which recently announced a $6 billion Series B funding round. The startup aims to "understand the true nature of the universe" and has released a chatbot named Grok, inspired by "The Hitchhiker's Guide to the Galaxy." Representatives for Musk and Altman did not immediately respond to requests for comment on the lawsuit's dismissal.

AI is poised to drive 160% increase in data center power demand

AI is set to significantly increase data center power demand by 160% by 2030, according to Goldman Sachs Research. Data centers, which currently consume 1-2% of global power, could see this rise to 3-4% by decade's end, driven by AI's extensive energy needs. A ChatGPT query, for example, uses nearly 10 times the electricity of a Google search. This increase in demand could lead to data centers accounting for 19% of their power use by 2028, contributing to a potential doubling of carbon dioxide emissions and necessitating substantial investment in renewable energy and emerging nuclear technologies.

In the US, power demand has been stable for a decade but is projected to surge by 2.4% from 2022 to 2030, with data centers responsible for about 0.9 percentage points of this growth. Data centers will use 8% of US power by 2030, compared to 3% in 2022, requiring $50 billion in new generation capacity. Additionally, increased natural gas demand will necessitate new pipeline infrastructure.

Europe faces even greater challenges, needing over €1.6 trillion to upgrade its power grid and support data centers. Power demand in Europe, which has declined by 10% since 2008, is expected to grow by 40-50% between 2023 and 2033. The power needs of European data centers by 2030 will match the current consumption of Portugal, Greece, and the Netherlands combined. Investments will focus on transmission, distribution, and renewable energy to support this growth.

Other stuff

FTC Chair Lina Khan shares how the agency is looking at AI
Why Apple is taking a small-model approach to generative AI
Brazil hires OpenAI to cut costs of court battles
Oracle shares jump on Google and OpenAI deals despite earnings miss
ARC Prize, a $1,000,000+ prize pool competition to beat and open-source a solution to the ARC-AGI eval.
OpenAI ex-employees worry about the company’s control over their millions of dollars in shares
Microsoft is killing off GPT Builder in Copilot Pro for consumers, just three months after broad availability
Spam, junk … slop? The latest wave of AI behind the ‘zombie internet’