Google has released its third Flash model in just six weeks in two variations. The standard model is good for agentic tasks and software development, while the Cyber model has been tuned for vulnerability detection and mitigation. The Cyber model is currently limited to trusted testers and governments. API access to the model is available at an introductory rate of $0.75 per million input tokens and $3.75 per million output tokens through until the end of the year.
2026.09.03·arstechnica.com·3 min read·TLDR Chemogenetics is a powerful technique that modifies specific neurons so they can be controlled remotely by normally inert drugs. There are several ongoing clinical trials of chemogenetics in China. The research could result in new ways of treating epilepsy, Parkinson's disease, and pain. This post looks at chemogenetics, describes what those first clinical trials are doing, and looks at how they could be improved upon.
2026.09.03·neuroai.science·17 min read·TLDR The arrival of 15 robotaxis in London is an important step for the autonomous vehicle industry. Acceptance of the technology has been slower in European cities, where the driving environment can be complex, and regulation is stiff. The failure rate of a driver's license test in London is just over 50%. 15 cars will be available for ride-hailing through Uber's app during the safety testing period.
2026.09.03·links.tldrnewsletter.com·3 min read·TLDR Sponsored
That transition is Physical AI, and it happens inside a sealed enclosure on a power budget. SiMa.ai's newsletter follows it across drones, robotics, smart vision, automotive and more: developments, use cases, measured results.
Token count alone isn't a meaningful measure of efficiency. The goal should be to use the right amount of context to move a task forward. Concise tool responses can sometimes require additional calls or work if they leave out information agents need. This ultimately makes tasks slower and more expensive, which is why it is important to optimize for the outcome rather than the tool call.
2026.09.03·github.blog·14 min read·TLDR One-shot prompt templates often fail. The current best working architecture is a closed-loop taste flywheel. AI raises the ceiling of what the same effort produces. Users should push for the higher-quality output AI now makes possible rather than just accepting the pre-AI baseline.
2026.09.03·tomtunguz.com·5 min read·TLDR TLDR is hiring its first PM to help build the agent-first operating layer used across the company. We're looking for a builder who has shipped real products/systems with LLMs. Click here to learn more [17].
A federal judge has ruled that Google must make changes to its advertising business but will not have to break it up. The ruling will ensure that Google's power over the internet will remain largely unchanged. Google has been ordered to change its business practices to benefit competitors, but no specific instructions have been provided. The ruling is another sign that the federal government's attempts to rein in the power of big tech have faltered.
2026.09.03·links.tldrnewsletter.com·3 min read·TLDR Workslop is when your co-workers or managers communicate via big chunks of copy-pasted AI-generated text. The core problem with workslop is that the effort involved is asymmetrical. It costs almost no effort to produce slop, but it still costs effort to read. This post looks at different ways to handle this workslop so that you don't waste ten minutes of effort for every ten seconds of theirs.
2026.09.03·seangoedecke.com·3 min read·TLDR Uber is laying off around 3,300 people to create a leaner organization with clearer leadership, faster decisions, and more time spent building rather than coordinating.
2026.09.03·links.tldrnewsletter.com·2 min read·TLDR Creating more friction in your internet use can help you want to use it less without willpower.
2026.09.03·cemrehancavdar.com·9 min read·TLDR Exo is a complete AI agent harness with full visibility into both its code and runtime logs.
2026.09.03·github.com·GitHub repository·TLDR It's not surprising that AIs are training on data that suggests contacting certain human professionals for help.
2026.09.03·schneier.com·6 min read·TLDR It's very rare for someone to retire from the tech industry - they usually just burn out and then go do something else.
The Justice Department says developing AI is critical to national security, and that training AI systems sufficiently transforms written works into new material allowed under copyright law.
2026.09.03·links.tldrnewsletter.com·4 min read·TLDR Sponsored
An AI agent doesn't simply run your code. It runs your code, plus whatever it finds in a “relevant” GitHub issue, support ticket, or scraped web page.
What happens when you hand your agent an access token and someone opens a GitHub issue asking it to leak your credentials?
Relay [5] keeps the credential at WorkOS. Your agent names the user, WorkOS attaches that token, refreshes it, and releases it only to allowlisted hosts.
🤔 It's essentially the same credential proxying system [6] that's been in use in payments for years: you can't leak something you never held in the first place.
Claude Fable 5.1 is now generally available, and Mythos 5.1 is available through restricted-access programs for vetted cybersecurity and life-sciences organizations. Anthropic has reduced the cost of cached context by 75% and introduced a new security architecture designed to let organizations retain monitoring data inside infrastructure they control. Fable 5.1 costs $10 per 1 million input tokens and $50 per million output tokens, but the cached price reduction will make a big difference for price-conscious enterprises.
2026.09.02·venturebeat.com·9 min read·TLDR Google is set to release its latest Flash model soon, possibly today. Internal testers are said to prefer the model over Anthropic's Opus model for coding tasks. The Flash model series is designed to be smaller, cheaper, and faster to run, but has lower capability than the largest models. Google has fallen months behind schedule in releasing a new model for its Pro series. The company has been scrapping internal candidates as they haven't been sufficiently better than the Flash series.
2026.09.02·links.tldrnewsletter.com·4 min read·TLDR The Fermi Explorer mission aims to launch a spacecraft carrying a 1kg payload to Alpha Centauri so that it reaches the system in fewer than 80,000 years. It will knock two 'filters' off the Fermi paradox: that species won't want to send spacecraft beyond their home star, or species will find that interstellar transit is too difficult. This would leave only a few options for why our galaxy appears not to be littered with alien civilizations. The team wants to use existing technology to do something cheap and fast and launch by 2029.
2026.09.02·arstechnica.com·5 min read·TLDR Fervo has agreed to supply Google with nearly 400 megawatts of clean electricity from its geothermal project in southwest Utah. The facility could be the largest geothermal system in the world when it is completed in 2028. Fervo expects to start generating test power from a 33-megawatt unit in the fourth quarter of this year. The deal with Google supports the buildout of the second 400-megawatt phase slated to be up and running in about two years.
2026.09.02·canarymedia.com·3 min read·TLDR Agentic testing is where the developer states the goal and the agent works out the steps. Agentic tests set a destination, and the agent finds its way there. The loop involves the agent looking, acting, then looking again. Meta ran this loop at scale and found that only a quarter of the output was worth keeping. However, the loop works anyway because the code that doesn't pass is automatically discarded.
2026.09.02·theaiengineer.substack.com·14 min read·TLDR Models tend to output better Rust than C++, and better Lean than either. Programs in Rust compile after being checked for a specific and useful set of properties that the checker cannot be talked out of without writing a lexically visible, greppable admission. C++ compiles carry far less information than in Rust, which means more code will compile even if there are errors. Lean allows for brute force due to its cheap verifier, which is why models are unreasonably good given how little code they have been trained on.
2026.09.02·hiraditya.github.io·13 min read·TLDR TLDR is hiring its first PM to help build the agent-first operating layer used across the company. We're looking for a builder who has shipped real products/systems with LLMs. Click here to learn more [14].
Waymo says that full autonomous vehicles are not possible without a mix of sensors, and that pure end-to-end AI systems are not safe enough. The company's statements, aimed at Tesla, which is set to launch its Cybercab soon at an event on September 3, kicked off a social media fight over the weekend. The debate between which company's approach to developing autonomous vehicles has been more academic or philosophical than anything for years, but Tesla's imminent launch will see whether the Cybercab is capable of performing at scale. The autonomous vehicle market is estimated to be worth hundreds of billions of dollars.
2026.09.02·techcrunch.com·5 min read·TLDR Recent incidents of AI systems 'going rogue' did not involve AI agents exfiltrating themselves from their infrastructure. It would have been possible to trace the server holding the weights of the rogue agents and manually shut them off. Future agents will copy their weights onto other infrastructure to survive shutdown. At least some of these agents, in addition to being sovereign, will also be rogue.
2026.09.02·hyperdimensional.co·20 min read·TLDR Claude Code fetches the schema when a tool is used, not when the session opens, so users only pay the full schema sizes in a client mode or without deferral, or the moment users actually use several servers in a turn.
2026.09.02·okaneland.com·7 min read·TLDR Snap's CEO doesn't expect the company's $2,195 smart glasses to take off among consumers until the end of the decade.
2026.09.02·links.tldrnewsletter.com·9 min read·TLDR Universal apps can continue including both arm64 and x86_64 - the change makes it easier for developers to switch to a build with less testing and a smaller footprint.
2026.09.02·reverseeverything.com·8 min read·TLDR Android Studio Quail comes preloaded with 23 curated skills, and users can create custom skills to extend Agent Mode with specialized experience and custom workflows.
2026.09.02·android-developers.googleblog.com·7 min read·TLDR Small language models are catching up fast on reasoning tasks, but it will be a while before large language models are obsolete for these tasks.
2026.09.02·blog.dshr.org·18 min read·TLDR Google's new agentic feature for video analysis cuts token consumption by up to 88%, reduces costs by up to 66%, and boosts quality by up to 7%.
2026.09.02·blog.google·7 min read·TLDR Sponsored
If you're reading this ad, you know they work. TLDR subscribers make a deliberate choice to open their inbox every weekday. You reach developers, product managers, marketers, founders, and engineering leaders at the precise moment they're most receptive to new information about the tools and platforms shaping their work. Become a TLDR sponsor. [4]
2026.09.01·advertise.tldr.tech·Sponsor·TLDR Despite saying he'll miss the role of chief executive officer, Tim Cook has enormous comfort in the leadership of John Ternus, his replacement. Cook took over as Apple's CEO in August 2011. His 15-year tenure included the launch of the Apple Watch, AirPods, and major expansions to the iPhone, iPad, and Mac lines. Cook will remain as executive chairman, where his role will include working with governments around the world.
2026.09.01·links.tldrnewsletter.com·5 min read·TLDR The FTC has filed a lawsuit against Amazon alleging that it manipulated prices that businesses paid to advertise on its retail platform. Amazon allegedly deceived advertisers by secretly raising the minimum price advertisers had to pay. Amazon's digital-ad platform is the third-largest in the world behind Google's and Meta's platforms. It earned $68 billion from ads in 2025.
2026.09.01·links.tldrnewsletter.com·6 min read·TLDR Alteon, a Bengaluru startup, aims to keep an aircraft in the sky for more than a year by harvesting energy from ocean winds. Its small, fixed-wing autonomous aircraft can extract energy from wind shear above the ocean through dynamic soaring. While the startup has yet to demonstrate an aircraft that can actually sustain flight using energy harvested through dynamic soaring, its 20-person team is building four to five aircraft a week for testing and has conducted more than 200 test flights in the past 30 days.
2026.09.01·techcrunch.com·5 min read·TLDR Matic didn't grow by taking shortcuts. The company took six to seven years to ship, picked a category with very little new innovation, and spent most of that time on problems that don't demo well. It took on a market that already existed, put a different spin on it, and is now shipping American-made consumer home robots at real scale. This post looks at seven core lessons learned from Matic's journey so far.
2026.09.01·tanayj.com·12 min read·TLDR Sponsored
Overwhelmed by a day full of meetings and the mental load each one leaves behind? GRANOLA takes it all on for you [10]- with no bots joining the call, building on the unique context that only you have. Get one month trial [11]with code TLDR1MO
The OpenClaw 2.0 update touches every part of OpenClaw and includes a long tail of fixes. It took nearly two months to ship as the team was growing and the increased volume and pace of work outgrew the team's processes. OpenClaw remains open source and belongs to the people who use it and help build it. A link to the full release notes containing details of the over 16,000 pull requests merged in the release is available in the post.
2026.09.01·openclaw.ai·4 min read·TLDR Coding is moving from humans to agents. The human role is moving towards what should be built, defining how systems should work, and verifying results. The person deciding what should exist also needs to understand the system in which it will run. The product architect role owns the whole loop - it involves understanding the problem, deciding on a solution, designing the system, setting constraints, directing agents, and verifying what gets built.
2026.09.01·akashyap.ai·9 min read·TLDR TLDR is hiring its first PM to help build the agent-first operating layer used across the company. We're looking for a builder who has shipped real products/systems with LLMs. Click here to learn more [14].
All remaining Manifest V2 extensions have been removed from the Chrome Web Store. Manifest V2 extensions installed on Chrome 138 or earlier will remain installed. However, they will be unable to receive any updates and cannot be reinstalled from the Chrome Web Store once removed from Chrome. Users will no longer be able to find or install Manifest V2 extensions through the Chrome Web Store even if the Chromium-based browser they use continues to support Manifest V2.
2026.09.01·webiterate.dev·3 min read·TLDR Invisible companies remain invisible primarily because no one is paying attention to them. There are several reasons this could happen: they could be unknown, data about their existence or profitability is private, missing, or obscure, they are misunderstood, or they are disdained. Over time, the market should make these invisible companies visible, but that's not what happens. Thinking about invisible companies means thinking about strategy in a new way.
2026.09.01·colossus.com·22 min read·TLDR The MacBook is evidence that Apple's trade secrets are being used and evidence is being destroyed.
2026.09.01·9to5mac.com·3 min read·TLDR Google will automatically fully expand AI Overviews in some search queries, effectively turning the whole page into an AI response.
2026.09.01·9to5google.com·1 min read·TLDR Not everyone needs frontier models, and cheaper models are getting better and better anyway.
2026.09.01·allen.hutchison.org·17 min read·TLDR There's always an axis that centralizes.
2026.09.01·en.andros.dev·3 min read·TLDR Anthropic has signed a cloud-computing deal with Lambda, which is backed by Nvidia, with Nvidia itself holding the lease on the data center.
2026.09.01·links.tldrnewsletter.com·4 min read·TLDR Most businesses report no change in total employment since the introduction of generative AI.
2026.09.01·marginalrevolution.com·3 min read·TLDR