TLDR 2026-08-17

Someone goes over budget. A receipt disappears. You lose time figuring out who spent what and why.

Mercury Spend replaces that system. It's expense management built directly into Mercury* — cards, budgets, and reimbursements in the same place as your business banking. Set spending limits per person, team, or card and let the policies enforce themselves. Receipts scan in automatically from Gmail, or arrive via photo upload or text. Cards with overdue tasks auto-lock until they're resolved.

No chasing. No manual review. No end-of-month scramble.

Explore Mercury Spend

Stripe has finalized an agreement to acquire OpenRouter, a startup that helps companies switch between AI models. The sale price of more than $7 billion could change and the discussions are not public. OpenRouter's rise highlights the industry's growing scrutiny on AI costs. The acquisition will give Stripe a stronger footing in the fast-growing AI sector.

Anthropic says it will not release an internal model it claims to be more powerful than Mythos and that it will not slow development broadly. The company believes that the risks of the most serious harms from its models are still low. While the company is seeing signs of acceleration in its models' ability to conduct automated research and development, it appears to be signaling that it is getting harder to understand the capabilities and risks of its own models. OpenAI says it is slowing down the release of an upcoming model called Astra because it can't rule out critical cyber capabilities.

Far-UVC 222 has been shown to efficiently inactivate airborne pathogens in a room-sized chamber. While the technology is unlikely to prevent the average common cold, mostly spread through close, extended contact, it could be helpful in blunting a future respiratory pandemic where more of the spread happens at a range through shared air. It is functionally equivalent to an extremely strong air purifier. The biggest bottleneck to the technology's adoption currently is awareness.

Heart Aerospace's X1 electric aircraft has a 106-foot wingspan and weighs more than 25,000 pounds. It flew for 27 minutes on battery power alone and reached an altitude of 1,100 feet during its recent maiden flight. Heart plans to bring a hybrid-electric plane into commercial service by 2031. The company's technology promises lower maintenance costs and insulation from volatile jet fuel prices for airlines.

Ads get ignored on social media but not in TLDR! Reach developers, PMs, marketers, founders and other tech leaders where they actually pay attention. Learn more about sponsorship opportunities.

If your most precious commodity is attention, ask your agents for the design after they've built the thing. Let the agent spend extra cycles in the lab, and get one layer deeper on the design. A plan that's been prototyped is a better plan.

The idea of a zero-knowledge proof (ZKP) is that there are two parties: the prover and the verifier. The prover asserts that it has a solution to a (generally NP-complete) problem. The prover can convince the verifier of this without sharing the actual solution to the problem. While they are used to secure some cryptocurrencies, that is not their only use case. This article looks at how ZKPs work and how they can be applied to other problems like age verification.

Several companies are attempting to build the synthetic superstars of tomorrow. AI imaging tools have enabled anyone with a modest budget and a sliver of vision to build their own photorealistic avatars. The availability of virtual talent has unsettled many communities of human performers. While AI influencers likely won't replace human influencers fully, they'll likely become more prevalent as these tools become more and more reliable.

TLDR is hiring a GTM Engineer to join our Applied AI team and own our AI-native GTM stack. We're looking for someone comfortable building AI agents and working with HubSpot. Click here to learn more!

Anthropic employee Jarred Sumner used the Claude app on his phone to try to solve the infamous Riemann hypothesis. The model didn't succeed, but it made a related finding that one Stanford number theorist has called the most impressive result that AI has produced in math so far. Sumner's formal mathematical education ended after just one semester of high-school geometry, and he identifies as very much not a mathematician. Most of the prompting he gave to the model was variations of 'keep going' and 'believe in yourself'.

OpenAI still can't define "AI agent." Business owners skipped the debate and hired Viktor, an AI employee in Slack & Teams. Dashboards. Apps. Tasks... Start free →

Z.ai released GLM-5.3 on Friday while holding back the model's downloadable weights and gating its most sensitive cybersecurity functions.

In a complex system, you can basically find as many bugs as you ask the agents for.

A machine can nail the task but completely miss the job.

The benchmarks are answering the wrong question.

This study introduces three tasks for AI agents that allow researchers to evaluate efficiency, collaboration, and taste, providing a proxy for how well the agents perform as tools in real coding situations.

Code is the byproduct of understanding.