AI Just Solved a Famous Math Problem

FEATURED INSIGHT💡

AI Just Crossed a New Line in Mathematics

For nearly 80 years, mathematicians have studied a deceptively simple question: if you place n points on a plane, how many pairs can be exactly one unit apart? This is the planar unit distance problem, first posed by Paul Erdős in 1946, and many believed the best possible answers would look roughly like square grids. Now, an internal OpenAI model has reportedly disproved that long-standing belief with a new family of constructions that performs better.

What makes this especially notable is that the result came from a general-purpose reasoning model, not a tool built specifically for this problem. The proof was reviewed by external mathematicians, and the solution connects discrete geometry with deep ideas from algebraic number theory, including algebraic number fields and class field towers. In plain English: the model found a surprising bridge between two areas of math that did not obviously belong together.

Why it matters: this is a real signal that AI is moving from useful assistant toward research partner. Not replacing experts, but helping explore difficult paths, connect distant ideas, and surface approaches humans may not have prioritized. If models can hold together long arguments that survive expert scrutiny, the same kind of reasoning could eventually accelerate work in biology, physics, engineering, medicine, and AI research itself.

Attio is the AI CRM for high-growth teams.

Connect your email, calls, product data and more, and Attio instantly builds your CRM with enriched data and complete context. Whether you’re running product-led growth or enterprise sales, Attio adapts to your unique GTM motion.

Then Ask Attio to plan your next move.

Run deep web research on prospects. Update your pipeline as you work. Find customers and draft outreach emails. Powered by Universal Context, Attio's intelligence layer, Attio searches, updates, and creates across your data to accelerate your workflow.

Ask more from your CRM.

Ask Attio

ON THE HORIZON 🌅

Quantum Computing Gets a $2B Government Push

The U.S. government is preparing to invest $2 billion into several quantum computing companies in exchange for equity stakes, including IBM, D-Wave, Rigetti Computing, Infleqtion, and GlobalFoundries. IBM is expected to receive $1 billion, which it would match to help create a dedicated quantum chip manufacturing company in Albany, New York.

The big idea: quantum computers use qubits, which can represent more than a simple 0 or 1. That could eventually make them powerful tools for cybersecurity, drug discovery, engineering, defense, materials science, and optimization problems that classical computers struggle to solve.

The catch is that quantum is still early. Today’s systems are fragile, error-prone, and not broadly practical for commercial use. Nvidia CEO Jensen Huang has suggested useful applications may still be about 20 years away, while Bill Gates has floated a much shorter timeline. Either way, the signal is clear: quantum is starting to look less like a science project and more like strategic infrastructure.

LATEST IMPORTANT NEWS 📰

Google signs a massive compute deal with SpaceX

Google will reportedly pay SpaceX $920 million per month from October 2026 through June 2029 for access to around 110,000 NVIDIA GPUs plus related compute infrastructure. Google framed the deal as short-term bridge capacity to meet unexpectedly high demand for Gemini Enterprise and its agent platform. The move also highlights how extreme the AI compute race has become: even one of the world’s largest AI infrastructure players is renting massive external capacity.

Apple approves Poke as its first AI agent on Messages for Business

Poke, an AI agent users interact with through text messages, has become the first AI agent approved for Apple’s Messages for Business platform. The startup helps users with planning, calendar management, fitness tracking, smart home controls, and photo editing through conversational messaging. The bigger story is distribution: AI agents may increasingly show up inside familiar communication channels instead of requiring users to download a new app or learn a complex interface.

Meta launches an AI creator assistant on Facebook

Meta is rolling out a conversational AI assistant that gives creators personalized recommendations based on their content, performance, audience, and goals. Creators can ask questions like when to post, what people are saying in the comments, or how their audience has changed over time. Meta is also expanding AI-translated Reels, with more than half a billion users reportedly watching AI-translated videos on Facebook each week.

Want to get the most out of ChatGPT?

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

Download the free guide.

FOR THE TECHNICALLY INCLINED 🛠️

A New Way to Make Diffusion Models Faster

A new paper introduces Strong Stochastic Flow Maps, a framework designed to make diffusion and flow models faster at inference while staying closer to the actual stochastic path of the underlying process.

The issue: many diffusion models require lots of network evaluations during sampling because they numerically integrate a differential equation. Flow maps try to shortcut that by learning the solution map directly, enabling few-step sampling. But earlier methods mainly handled deterministic ODE-style paths or recovered only the right marginal distributions of an SDE. SSFMs aim for strong convergence, meaning they learn the actual solution path for additive-noise stochastic differential equations. The authors also introduce a polynomial approximation to Brownian motion and show simulation-free training for diffusion model solution maps, with results in image generation and few-step molecular system sampling.

Translation: fewer sampling steps, better path fidelity, and potential speedups for both generative media and scientific modeling.

AI TOOL OF THE DAY 🚀

Orange Slice helps automate sales workflows by finding customers, enriching lead data, and building AI-powered revenue workflows that can handle real-world complexity.

One AI employee. Engineering, finance, growth, ops.

Last week Viktor opened 14 pull requests, closed two month-end books, drafted a board update, deployed three landing pages, and triaged 600 support tickets. From inside Slack and Microsoft Teams. 20,000+ teams now run this way.

Get Started for Free