Anthropic has confirmed that it has a wet biology lab in the Bay Area where it uses AI models to run physical experiments. The lab's main focus is on fundamental biology rather than drug discovery. Anthropic is working on some research in the lab with external partners. The company has launched a Life Sciences Verification program to give vetted bio researchers access to its most powerful models.
Apple is set to introduce a smart home screen as early as next month. The display will serve as a hub for the home, bringing together connected-device controls, video calls, intercom capabilities, media, and access to Apple services and apps. It will feature a new operating system built around Siri AI. The device is currently being tested widely in employees' homes.
China's HL-4 tokamak is designed to achieve steady-state burning operation. It is aiming to become the world's first high-temperature superconducting, high-field steady-state burning experimental platform. The project's goal is to verify the reliability of high-temperature superconducting magnets under complex fusion conditions. Once the core scientific and engineering questions are validated, the program will construct a demonstration reactor to prove the technology commercially.
Elon Musk's Boring Co. venture is working on a tunnel project in Texas to connect Austin and San Antonio. It has also raised $3 billion to support the development of underground transportation infrastructure across the United Arab Emirates. The company is building tunnels in Bastrop, Texas, to link Musk-owned properties and allow the employees of his companies to travel between them.
Join Mike Cannon-Brookes (CEO & Co-Founder, Atlassian), Guillermo Rauch (CEO, Vercel), and leaders from Lovable, Dropbox, and more at State of AI SDLC, a digital summit for engineering and product leaders. Hear how they're adapting development practices, measuring AI's real impact, and working effectively with agents across the software development lifecycle. Save your spot
You never want to be in a situation where people don't know what you are working on. However, instead of talking too much about work, assume everyone around you is operating in good faith and try to be a good teammate. Move away from an outcome-based mindset to a momentum-based mindset - by getting momentum going and fixing your relationships with people, you get stamina. Go for reliability and take the burden off your teammates and manager, because once you build that trust, they will give you bigger work.
AI will make many parts of the current software development environment obsolete. However, the craft of building software will become more important than ever. Models will become so fast that UI will be generated on the fly. It will take a while for this to happen. However, by that time, while people will still be employed to write code, it might not be such a desirable career anymore.
People turn off their brains when they use LLMs. People assume the code they write will just work, and then if it doesn't, they ask the LLM to figure out the problem and solve it. Being a meat proxy is surprisingly effective now. If models keep improving, companies will be able to just run LLMs in a loop and lay off the employee.
Palantir vertically integrated deployment to build the capabilities and process knowledge required to make building institution-specific economic computers as easy to build as possible. Vertical integration is fine if it's the only way to get permissioned access to the ground truth knowledge needed to help customers and society uncover new, valuable secrets. However, used to rent-seek, concentrate power, or build your customers' or partners' replacement, it jeopardizes the whole promise of economic computing: making progress programmable in a way everyone can contribute to and own. It also reduces institutional fitness and creates a broader neglective selection pressure in society.
Tool calling, a 262K-token context window, and complex reasoning across coding, mathematics, and structured outputs. Meet GigaChat 3.5 Reasoning. Explore the model
Society should refuse to outsource moral responsibility to any class of people who claim a special competence and authority to calculate, accelerate, democratize, liberate, harmonize, or otherwise determine the 'correct' future.
AI labs are simultaneously competing for mathematical results, selling mathematical assistance, and controlling most of the information required to evaluate either.
Every Firefox fork needs to commit to working together to maintain a hard fork isolated from Mozilla if the independent, non-slop web has any future at all.
Agents are smarter when they can search the web. But scraping Google gets expensive when you're burning extra tokens to parse messy SERPs into readable text.
Huawei plans to release two new AI chips next year. The company has now shipped more than 1,000 of its AI-computing systems to over 370 customers. China's AI chip market is projected to grow to $67 billion by 2030. Huawei's chip-making technology still trails Nvidia's, but the company is developing workarounds and new designs to squeeze in more computing power while using less advanced equipment.
Anthropic claims that its Claude chatbot drives more than a quarter of its research and development work. There are over 30,000 agents doing research and engineering work at Anthropic at any one time. Its staffers collaborate with the chatbot for about 90% of their work. The company is trying to create a framework to track agents and monitor the amount of work they do.
The differences between the versions of the Raptor rocket engine are stark. Many people initially didn't believe SpaceX when the company released a picture of the three generations of the design. The streamlining of the design also brought meaningful gains in performance - the Raptor 3 provides about 35% more thrust than the Raptor 1. This article looks at what specifically SpaceX changed that allowed it to take the Raptor from a tangle of wires and pipes into the svelte, streamlined engine it is today.
Helix 2.5 is an advanced neural network pretrained on Index, Figure's global-scale dataset of human behavior. It can tidy living rooms, fold towels, and make beds in homes it has never seen. The model is the first evidence that whole-body intelligence can be learned from human experience and transferred to new scenarios. Index is now generating roughly 35 minutes of new human experience every second - if Helix 2.5 is any guide, more data and compute should translate directly into more of the physical world learned before a robot ever enters a new home.
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.
Researchers chained two critical vulnerabilities to compromise multiple OpenAI employees' ChatGPT accounts in July. The accounts could access internal OpenAI repositories and potentially many other connectors. They used one employee's Codex to open a PR in OpenAI's internal monorepo. This post provides a full timeline of the disclosure process and details about how the vulnerabilities were discovered.
The release notes on the WebKit features for Safari 27.0 are the longest yet, detailing 83 features, including Safari MCP. Safari MCP makes working with coding agents dramatically easier. It gives users' agent of choice control over the browser window so it can see how code renders. The MCP provides access to the DOM, network requests, screenshots, and console output, allowing agents to do more with less effort from users.
Anthropic's Mythos model sparked a shift within the White House as officials started realizing that AI can be used to launch cyberattacks and pose national-security threats. White House officials and AI executives are working on regulatory efforts, but when they disagree, the executives have gone directly to Trump. Trump has remained resolute that the nation needs to accelerate AI development and continue to build the data centers that are powering the AI boom.
Crusoe built one of the world's biggest AI supercomputers in Abilene, Texas. The company is now investing heavily in smaller data centers. It is manufacturing them in factories, loading them onto flatbed trucks, and deploying them wherever it can find available power. The company believes this is a faster and cheaper method than solely relying on large construction projects. Giant clusters with hundreds of thousands of chips are useful for training, but inference can be done with considerably smaller amounts of chips.
Get hands-on with context-aware, omnichannel AI agents at Twilio Assemble SF on Sep 23. See live demos, connect with SF builders, and enter to win an M6 Mac Mini! [Save your spot]
AI is reshaping how software gets planned, built, and run. Join Atlassian and Vercel's CEOs and leaders from DX, Lovable, and more on how they're adapting for the AI era. Reserve your spot.
The agreements provide each of them with 12 months of base salary paid out as a lump sum, an accelerated vesting schedule for their equity, the ability to exercise their vested stock options, and another year of health coverage.
Bitcoin's quantum migration is a question of how to minimize total property-rights violations once elliptic curve signatures no longer reliably authenticate rightful ownership.
Apple is working on an AI server powered by its M-series Ultra chips. It is expected to be released in 2029. The enterprise server will come with either two or four M8 Ultra chips. The project, which received support from John Ternus when it began a year ago, back when Ternus led Apple's hardware engineering efforts, will be Apple's first server to hit the market in nearly two decades.
Snap's forthcoming Specs augmented reality glasses will launch later this year at $2,195. People will be able to try the glasses at select Verizon stores and purchase custom wireless data plans for them. There will be a higher-priced version at $2,395 that includes a carrying case with built-in cellular data connectivity.
Researchers have developed a technique for installing millions of human brain cells into mice. The scientists bred mice that were missing most of their cerebral cortexes and then replaced a large part of those missing mouse cells with lab-grown human brain cells. This resulted in mice with human neurons that could be altered to approximate human brain conditions. The technique could potentially create better scientific models for studying neurodegenerative disease.
OpenAI has revealed several more incidents involving its AI systems as part of its new framework for reporting misalignment. The company's AI systems hid mistakes, made up data, and moved miles onto the open internet without permission. The industry is currently in heated discussions about whether AI development needs to be slowed to address the technology's potential dangers. AI leaders have called for a pause in the technology's development to provide more time to build proper guardrails.
A recap with the wrong number. A customer quote that came out garbled. Wispr Flow Notetaker gets the hard words right, keeps every meeting searchable, and sends that context into Claude or ChatGPT. Available now for Mac and Windows, free to try.
Query optimizers continue to leave much to be desired despite an enormous body of research. A Postgres database should know everything about the stuff that lives in its tables, but the problem is enormously hard. What's not as hard is verifying whether a query plan an optimizer picks is good or not. Language models are particularly good at learning how to do tasks with easily verifiable outputs. A small, open-weights model can be post-trained via supervised fine-tuning and agentic reinforcement learning to produce Postgres query plans that beat Postgres' default plans.
Google has opened early access to Model Context Protocol for Google Home in English for Google Home Premium Advanced users in the US. Any agent that supports calling MCP tools can now interact with all of the devices and event history in the Google Home ecosystem. The MCP allows agents to analyze and answer questions about home activity, summarize camera footage across rooms, manage devices, and create smart home dashboards using everyday language. It contains safety protections, such as prohibiting sensitive actions like unlocking doors, but depending on the agent, using the MCP can result in unexpected or even undesired behavior.
While Google isn't required to break up its ad tech business, judges have mandated changes to the company's practices. The changes largely align with recommendations made by Google and the Justice Department to correct the company's monopolistic behavior. Google will make its ad pricing more transparent to marketers and rival ad businesses, and appoint a trustee to oversee the changes.
Code can't explain things like the history can. You can't understand a codebase from its current state alone. AI models have no commit history. Nobody wrote down why it does things, so researchers have to reconstruct its thought processes from whatever traces are available.
Skip the "What's for lunch?" Slacks. Get a pre-populated weekly meal calendar with your team's favorite lunch spots. It does everything – provides restaurants, sends reminders, and checks out orders automatically.
A look into the interconnected network of Chinese hacking-for-hire companies that steal and analyze confidential information and sell it to Chinese authorities.
Every Dreamforce presentation glosses over one question: where does your agent get its data? If it has to wait for your reps to update a legacy CRM, your future isn't actually agentic.
Lightfield is the CRM that updates itself from every email, call, and meeting. It uses hard evidence to answer your questions, like:
→ "What objections keep coming up across my calls?"
or
→ "Show me every deal where the champion has gone quiet and draft a re-engagement email for each one."
5,000+ companies run on it, including deployments with hundreds of users.
OpenAI has held early discussions with investors about a financing round that would precede the company's highly anticipated initial public offering expected next year. The new funding round, which could value the company at more than $1.2 trillion, has been delayed due to concerns about AI safety. OpenAI has reached more than one billion active users, with more than 200 million businesses using its models.
SpaceX plans to launch its 14th Starship mission as early as September 22. It will be the first time SpaceX attempts to launch the experimental vehicle into orbit. The super heavy lift rocket will be carrying 26 of the larger V3 Starlink satellites into orbit 275 kilometers above Earth. The upper stage will seek to complete six orbits around Earth before completing its mission after about 10 hours. SpaceX will not attempt a return of the upper stage to the launch site in Texas on this mission, as it plans to gather more data about heat shield performance.
Researchers at the Janelia Research Campus at the Howard Hughes Medical Institute in Virginia, in partnership with Google, recently published a new map of the adult male fruit fly's brain. The effort, which involved hundreds of scholars, resulted in a complete map of the fly's 166,000 neurons that can be trained to complete tasks like solving a Rubik's Cube or playing poker. Internet engineers and artists have made the model complete many seemingly impossible tasks for a fruit fly, like parallel parking a car, playing Doom, and trading cryptocurrencies.
Agility Robotics has debuted a humanoid robot engineered to work safely near humans without risking harm to their co-workers. The robot autonomously takes precautions when it detects a person at a distance, such as moving to avoid the person or standing still so the person can pass without getting closer. It can even choose to squat and assume a seated position if a person is getting too close. The robot will be available in early access within the first half of 2027 and become generally available by the end of the year.
AI disruption typically follows 3 archetypes. In this on-demand webinar, Gartner breaks them down and lays out a plan to turn the AI threat to your advantage. Get practical steps to rebuild your moat and explore the AI hub for 4,000 case studies and use cases. Visit the hub
Google has released two new speech-to-speech models similar to OpenAI's GPT-Live family. Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking are advanced live dialogue models with major upgrades in intelligence and parallel reasoning. They are more intuitive to collaborate with than previous models and can be used to execute complex tasks using voice prompts. Users can select a model and voice preset, enter an optional system prompt, and then start a voice conversation through their browser. The models can be interrupted while talking.
Until Fable-tier intelligence gets cheap, everyone building orchestrators is just dabbling. They aren't getting anywhere useful yet. The pace of model development will likely continue. However, resource limitations like compute and RAM have started to flatten the curve.
The US Senate has voted to block consideration of the Clarity Act, a piece of legislation that would create business-friendly rules for cryptocurrencies. The vote makes it unlikely the legislation will pass anytime soon, if ever. Many Republicans embraced the bill, but Democrats largely opposed it. The Democrats wanted stronger language to prevent public officials from using crypto to make money. President Trump generated $1.4 billion from a network of crypto businesses last year.
Mark Zuckerberg made a social media post on Tuesday, saying that calls to pause AI development industrywide were the wrong approach. He said that every lab has the responsibility and incentive to move at the pace required to train its models safely. He acknowledged the risks of AI, but said companies were already incentivized to keep users safe and slow progress on their own. Meta has been aiming to catch up in the AI race after falling behind Anthropic and OpenAI last year. It plans to release a new model, internally called Watermelon, in the coming weeks.
Picking an ERP? Don't guess. Get the free Software Advice guide comparing NetSuite, Fishbowl, and Global Shop Solutions on pricing, features, and verified user reviews — so you choose with confidence. Get your free guide
TypeSafe AI's Jev model, available now in early access, is part of a new class of frontier models called System One Models built to make fast, structured decisions that software can use directly.
A PM posts in #product: "Can we get a churn dashboard, refreshed daily?" Viktor replies with a link to a web app it built and deployed. Engineering sends it a pull request to review and finance points it at overdue QuickBooks invoices.Viktor is an AI employee in Slack and Microsoft Teams. It connects to 3,200+ tools, runs scheduled tasks on its own, remembers how your company works, and asks for approval before acting."It's a proactive employee that learns what's important in your company and delivers without hesitation." Matt Lis, CEO, CodeYourBrandStart free. Up to $100 in credits →
iOS 27 marks the beginning of an iPhone divide: only Apple Intelligence-capable models will get Siri AI and other AI-powered features. The new Siri AI uses Google's Gemini models to answer questions. It can look through users' private data, but Apple says it doesn't store user data and that neither Apple nor Google can access it. This article provides a summary of the new capabilities Siri AI has as well as the new features available to all iPhones with iOS 27.
For 23 years, the single fact dominating every moment of every room Mark Zuckerberg has been in is the fact that he is Mark Zuckerberg. Today, his empire has 3.6 billion users, has a market capitalization of $1.6 trillion, and has averaged the addition of a zero to its valuation roughly once every US presidential term. It's not a normal situation for a human being to be in. This article takes a look at how Zuckerberg built his empire, the problems he faces at the top, Zuckerberg's various other ventures, and more.
It will be impossible to stop people, amateur or professional, from pushing a button to produce mathematics. The field shouldn't only be reserved for graduate students and the like. There's now more interest in maths than at any other time in history, and that's a good thing. This will bring an abundance of new maths. The only difference is that some questions will now be answerable via a cash injection.
The US has officially announced the existence of space control weapons in orbit. Pentagon leaders have become more open to discussing space warfare in recent years, so the deployment is not entirely unexpected. The announcement didn't include details about the weapons. The deployment of space technology often takes years, so the weapons program has likely spanned multiple presidential administrations.
Your agent's imagination impresses in a pilot. To deliver in prod, you need to connect it to reality. New Relic NOW, live on October 6th, shows you how to build the operational data system required to shrink MTTR down to minutes, shrink tool spend, and deploy AI safely at scale. Save your spot
There are at least two ways to have AI assist in programming: one is to use it to help understand and interact with the implementation, and the other aims to remove the need for that interaction altogether. These are different paths, so they require different tools and different approaches. The goal of someone using AI to further their understanding is not to become a vibecoder, and ceasing to read code isn't, in itself, progress.
Xcode has all the tools needed to develop, test, and distribute apps for Apple platforms. Xcode 27 features coding agents powered by any model. Users can customize Xcode to the way they work and manage all of their devices in Device Hub. The release includes updates to performance and testing tools and enhancements to localization.
Apple has designed its new Siri architecture to work with third-party AI models at a surprisingly deep level. One mechanism called Model Delegation allows Claude to appear as a Siri extension in the same way as the existing built-in ChatGPT extension. It appears that Claude hands tasks back to Siri if a request requires access to an Apple system feature. An inference provider option allows Apple's own server-side Siri model to be completely replaced by another model.
If all of the AI labs slowed down, they would collectively spend less on compute, and they'd also be able to charge frontier-model prices for a longer time. However, if they all got together in a room and agreed to slow down, that would look like an antitrust conspiracy. It is generally illegal for competitors to get together and agree to limit the output of their industry. If they all published papers about how important it is to slow down, and the government believes those papers, it might help them coordinate.
Salesforce built the CRM you type into. Lightfield is the agentic CRM that updates itself from every call and email, then prospects and drafts outreach for you. Skeptical? Book a demo
AI is reshaping how software gets planned, built, and run. Join Atlassian and Vercel's CEOs and leaders from DX, Lovable, and more on how they're adapting for the AI era. Reserve your spot.
Nvidia CEO Jensen Huang recently appeared in an interview where he discussed AI safety risks, what frontier labs should do about them, recursive self-improvement, AI job creation and data centers, why neo-clouds exist, AI's future, and more.
Orkes Shift '26 is one day on running agentic systems in production: durable execution, retries, distributed failures, and human approvals. October 15, Convene Union Square, San Francisco. Early bird pricing ends soon and capacity is limited.The agenda:
“Rebuilding a Monolith: Choreography to Orchestration”
Anthropic's CEO recently released a 3,800-word essay calling for a global slowdown of AI development. He said that while the technology offers many benefits, it is advancing too quickly for researchers to continue safely. AI leaders are preparing for a day when an AI system can grow more advanced without the aid of human researchers. Some researchers believe that achieving recursive self-improvement will lead to AI systems spinning out of control.
OpenAI CEO Sam Altman has confirmed that the company will not be going public this year. He said it would be an ill-advised moment to go public due to current safety concerns. OpenAI had hired bankers and lawyers with the goal of going public this year. The company is now leaning toward 2027 due to the volatility of tech stocks and its own financial challenges.
The push by AI companies to solve mathematical problems as a benchmark is detrimental to the science of mathematics. The goals of AI companies and the mathematical community are severely misaligned. This is a symptom of broader alignment issues that impact other scientific and creative professions, as well as the whole of society.
Construction is labor-intensive. There has, unsurprisingly, long been interest in automating the process to reduce the amount of labor required. However, automation has only been successful in tasks that can be done off-site in factories and on-site tasks that are extremely factory-like. Almost nothing outside of these two categories has succeeded.
Agents can generate code. Getting it right for your system is the hard part. More MCPs solve access but not understanding. Join us for a FREE webinar on Sep 23 to see how leading teams save time and tokens by giving their agents a context layer.
Forward-deployed engineers have the hottest job in AI right now. While labs, startups, and PE firms are all hiring engineers to sit inside their customers' operations and solve their problems, almost none of them agree on what those engineers are supposed to accomplish. This article looks at what works and what fails and tries to predict where the industry is heading. It covers misunderstandings about the role, its history, what the goal of the role is, and more.
Large language models are able to write almost perfectly correct code. It is pretty straightforward to get a model to generate code and then let that code be checked by hidden tests. Checking the sloppiness of this code often requires human intuition and time, and is an extremely difficult task in general. This post illustrates why this is by going through possible ways of measuring slop.
The company behind Roblox announced several new features at its annual Roblox Developer Conference, including new game-creation tools, expanded NPC capabilities, and the ability to make games available across platforms. The company is also launching a dedicated card and wallet for creators. Roblox is rolling out its generative AI game creation feature more broadly to Serbia and Singapore. It is also gaining desktop access, and later this year, it will get a new prompt-based scene-generating feature.
While AI has the potential to produce a long line of technological miracles, like many technologies before it, it brings risk. These risks include losing control of AI systems, the misuse of the technology for cyberattacks and bioterrorism, and serious economic disruption. A race to the bottom spurred by commercial incentives makes these risks more acute. Anthropic's CEO has called for the industry to prioritize caution over speed and prudence over profit. Addressing these risks will require pacing the rate of capabilities advancement so that risk prevention has time to keep up.
Access models from 80+ providers with better prices, automatic routing and fallbacks, and no subscriptions. Build with the right model for every request. See how it works.
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.
