Devs finally have a challenger to GitHub code hosting with Origin.

Origin was built on Buildkite.

Buildkite is a launch partner for this new code ecosystem. Origin holds the code. Buildkite runs the build. Vercel ships it. Three layers, one path from idea to production.

We also run CI for OpenAI, Anthropic, Mistral, xAI, Airbnb, Nvidia, Canva, Uber, Meta, Pierre, Bun and more.

Build faster, more securely, more reliably, with the CI engineered for frontier scale, used by the frontier labs setting the pace.

Start with our 30-day all-access trial.Get building →

Neuralink's decoders are machine learning models trained to convert neural activity into intended action. These decoders are unique to each user - all users have to perform calibration prior to using a Neuralink device. Neuralink has collected over 50,000 hours of unlabeled, freeform neural data since its clinical trials began. This post details how the company used this data to improve its models, making its decoders more accurate and long-lived.

Apple CEO John Ternus is spending a lot of time overseeing the company's hardware and software design work. He is largely leaving the company's financial operations to CFO Kevan Parekh and the supply chain to COO Sabih Khan. Ternus is deeply involved in product engineering and design. One of his first priorities was to restore the company's design studio standing. The company is not looking to recruit an outside chief design officer to oversee the entire organization.

The series of Harvard studies that showed that nuclear power plants were associated with higher rates of cancer and cancer mortality are nonsense. The methodology used in the studies will always show an increased cancer risk and mortality no matter what landmark is used. The studies simply do not prove anything about cancer incidence. Sophisticated statistics can produce extremely precise nonsense when the design fails to identify the causal mechanism.

NASA wants contractors to prepare to build a nuclear reactor that could survive a space voyage and run without maintenance near the south pole by December 2030. The agency is racing for the future of space exploration. The US and China are both racing to establish a moon base to search for resources and to launch missions deeper into the solar system. Controlling access to the Moon's assets is vital to this mission. Both superpowers believe a nuclear reactor is the centerpiece of this ambition.

Agents can generate code. Getting it right for your system is the hard part. More MCPs solve access but not understanding. Join Unblocked live on Oct 7 to see how leading Eng teams save time and tokens with an organizational context layer.

e2e is an open-source AI testing framework. It provides customizable testing for web, mobile apps, and more. e2e remembers agents' actions and replays them when possible without calling the model again. Less repeated reasoning results in fewer tokens and lower costs. Developers can choose their preferred AI model and customize the testing agent.

Advocates for web standards, performance, and accessibility have implored web developers for years to 'use the platform', but despite how obvious the advice seems to be, many people seem to still need convincing. One reason for this is that, for the longest time, browsers were playing catch-up with the ecosystem on top of them. Another is due to the familiarity with tools they've used in the past. Often, even with a robust standard, libraries fill a useful gap between framework ergonomics and the platform underneath it.

David Robinson resigned from OpenAI this week, joining a parade of former colleagues who have decided that the company's current path is unacceptable. Robinson headed the writing of the safety reports that OpenAI publishes with each major launch. OpenAI has thrived by trial and error, but this approach guarantees periodic failures, and the scale of those failures is growing. Achieving something much closer to perfect the first time is becoming more essential because iteration after a mistake may not be possible.

Alexandr Wang, 29, was hand-picked by Mark Zuckerberg last year to oversee Meta's AI efforts. Wang, the company's first Gen Z senior executive, brings an internet-flavored communication style that could help the company connect with younger customers. He is a native of the social-media ecosystem his boss built. Wang has succeeded in building hype for Muse, in part with an unconventional internet campaign. The challenge is now to keep the momentum going as rival companies press ahead with their own offerings.

Baseten pairs custom kernels with cross-cloud capacity, so open-source and fine-tuned models scale in its cloud or your VPC. Start deploying →

Microsoft could win the enterprise harness with Copilot while labs compete to supply its intelligence.

The swarm chasers are a loose crew of online sleuths that work together to piece together interconnected AI swarm incidents.

Trades involve physical demands and uncertain rewards that people outside the industry just don't understand.

People who want to live dangerously should be able to do that, but it needs to be on an opt-in basis.

None of Muse's capabilities are new, Meta just put the pieces together in the right way.

Neko Health uses 'advanced technology' to screen for issues in the blood, skin, and metabolism.

80% of product professionals say AI helps them ship faster, but customers aren't seeing value any sooner. Atlassian's State of Product 2027 explores the gains AI is delivering and the challenges that remain, from decision-making that hasn't kept pace to gut instinct overriding customer evidence.

Read the report

Starlink plans to offer hourly, daily, and weekly passes for its satellite internet service. The company is recruiting 'hosts' to sell passes to interested consumers. Pricing details have yet to be released. Starlink will take care of payments, user access, and connectivity while the hosts focus on setting up their sites and earning from connections.

Meta is one of the very few companies that can monetize a personal agent through advertising without ever showing an ad inside the agent itself. The company can keep ads out of the Muse app and still get paid for it on Facebook and Instagram. Every Muse errand can be an incremental signal. Meta can indirectly monetize Muse until enough demand has been aggregated for it to start with direct monetization.

Fervo Energy has started selling electricity from its Cape Station geothermal power plant. It is the first enhanced geothermal company to reach a key commercial milestone. The entire site has the potential to generate as much as 4 gigawatts of electricity. The first block at the site took 23 months to complete - Fervo aims to complete future blocks in as little as 18 months.

NASA's Probe Explorers are a new line of observatories intended to do more science for less money. The project aims to seek a better balance between the smaller Explorer missions and multibillion-dollar flagships. NASA announced on Wednesday it was moving forward with the first of the Explorers, PRIMA (Probe far-Infrared Mission for Astrophysics). PRIMA will observe the universe in faint, far-infrared light in a range of wavelengths between 24 and 235 micrometers. NASA intends to launch the telescope in 2033.

Pi Durable is an experimental package built specifically for long-running, durable, and malleable agents that can run anywhere. It can run anywhere, can be reached from different surfaces, supports infinitely long conversations, survives catastrophic internal and external failures, and lets multiple humans steer the same agents. It is a framework for building any agentic application, including coding agents.

David Heinemeier Hansson, the creator of Ruby on Rails, recently announced that 37signals, a company known for its software craft, was done with writing code by hand. The maturation of AI tools has made it possible for the company to embrace a future where coding by hand is almost entirely absent. AI creates new business models and revenue streams and disrupts traditional software. Companies are now investing in innovation and are willing to pay top-of-market for software engineers who can help them build AI products or become AI-native.

Asha Sharma, the 38-year-old chief executive of Microsoft's Xbox division, wants Xbox properties to entertain a billion people. This means opening untapped markets in Africa, Latin America, and South Asia by using cloud computing technology to run games on computers and smartphones. Another pillar of her strategy is to reinvest in Minecraft as a competitor to Roblox. Leaders at the company describe the future of its consoles as embodying that idea.

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.

AI is still too limited to substitute for large amounts of human work. However, organizing agents is no longer the hard part. Agents may be easier to integrate into firms than expected, as long as humans are guiding them in the right direction. This could mean more work for people, not less.

16,000+ investors have put $45,000,000+ into Doroni, maker of the H1-X flying car, after a decade of R&D. First deliveries are targeted for 2028, in a market Morgan Stanley projects at $9T by 2050.

Invest at $3.30/share before 10/23

Clef and Clef-flash are fully Jev-API compatible decision models that help agents programmatically gather context, make decisions, and take actions on tasks.

Transcripts record what agents were asked and what they did, so watch them for early warning signs.

For stablecoins to become the new backbone of the financial system, the mix of volumes passing through them will have to converge to those of traditional finance.

The potential deal will help Anthropic and other companies access chips and other key infrastructure.

arXiv now limits submitters to up to two submissions per calendar month, with a limit of three total active submissions at any given time.

Amazon Aurora PostgreSQL users can now directly query operational data together with data stored in their data lakes in Apache Iceberg and Apache Parquet formats using existing PostgreSQL applications and tools.

Modulate raised $25M, led by Future Ventures, to build the understanding layer for voice AI.

Transcripts drop the fraud, frustration, and abuse hiding in conversations. Velma, Modulate's audio-native model, catches them by orchestrating 100+ voice-specialized submodels running on raw audio.

✅ Stop fraud→ catch deepfakes with the #1-ranked detection model on Hugging Face✅ Catch AI agent failures→ flag hallucinations and compliance violations in real time✅ Keep customers→ spot churn and dissatisfaction signals✅ Protect human and AI agents→ flag risk and abuse✅ Build your foundation with it→self-serve APIs; top ranked Hugging Face Leaderboard STT API

The funding brings more models, deployment options, and broadens the APIs for teams building with voice.

Start building with Modulate's API

Apple is preparing to launch several smart home devices on October 13. The company plans to ship a smart-home hub, an updated HomePod mini, and a new TV set-top box. The company is developing a broader smart-home ecosystem that expands on HomeKit software and supports new categories of connected products. It is also working with third-party manufacturers on devices designed for the Siri AI home platform.

Google's Gemini 4 Argon model offers industry-leading performance in coding, knowledge work, and cybersecurity. Google's engineers are already using the model extensively, and Argon-powered agents have been working to migrate C/C++ codebases to Rust across Google. The model supports an output limit of 1 million tokens. It will be priced at $2 per million input tokens and $10 per million output tokens for a limited time. The model is still in limited testing, and there is no timeline for when it will be available to enterprise and consumer customers.

The DeepMind team at Google has developed a technique that can create watermarks on protein sequences without compromising the protein's function. The watermark is impossible to remove and can survive basic exporting, resizing, and other editing. The system could be used to give DNA synthesizers greater confidence in their sources.

Late-stage clinical trial data on the obesity drug retatrutide showed an average loss of 25% of body weight after 90 weeks. More than a third of the participants taking the drug in the trial lost 30% or more of their weight. The drug also helped reduce knee pain and sleep apnea events, and it improved cardiometabolic measurements across the board. At the start of the trial, more than a third of trial participants had prediabetes - the condition had resolved in more than 90% of those participants by the end.

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.

Developers can now programmatically create, read, and manage comments across Google Docs, Sheets, and Slides. The Google Docs API also supports suggested edits. Organizations can now connect internal review tools, automated content pipelines, and project tracking systems directly into Google Workspace editors. Programmatic comment support is available to all Google Workspace customers and users with personal Google accounts.

Writing a commit description can help you reflect on the changes that AI made. It also serves as a way to check if everything is as intended. Being able to explain why you are shipping something will help you explain or fix it if it breaks later. AI can do the work - writing why is where you find out whether you understand what you are shipping.

AI chief executives gathered with Trump at the White House on Tuesday to discuss a statement of principles to keep AI models safe. Behind the scenes, the executives questioned Anthropic CEO Dario Amodei about his warnings about the dangers of AI, asking why he was being so extreme in public about the capabilities of AI models and the risks they pose. Amodei told them it was important to be honest with the public about model capabilities and not play down risks. He believes that safety is possible if the industry and administration work together.

AI assistants are becoming a router between people and every business. They decide who gets the work, which moves where money can be made. The router absorbs capability and decision-making. The margin goes to whoever holds what the router cannot take and whoever signs for what the router cannot sign.

Clarify replaces your CRM, outbound, enrichment, and call intelligence with one autonomous system—so you spend less time managing your sales stack and more time selling. Try Clarify free

300+ events worldwide to learn about open-source AI and open-weight models. Free Arduinos and T-shirts at select events. Find a Fest near you.

The new tool, which lets users send prompts like 'order my usual' to get their regular Friday night order, is available to users in the US through a waitlist.

The amount of borrowed money in the stock market has never been larger, or grown faster.

What you do matters less and less as fewer people know of you and more people know you.

RSS support on Reddit will cease on November 13 as the company says it has become a common surface for large-scale scraping and automated abuse.

Tech contributed around 76% of the SP500's total earnings growth in 2026.

There are several techniques for letting people compute something together without sharing the underlying information.

Span connects coding agent sessions to tickets, PRs, and production outcomes, giving you evidence to measure returns and identify what to improve.

>Measure results: Benchmark delivery, quality, and efficiency, then track progress against your targets.

>Direct investment: See where tokens and team time go, what they produce, and how that aligns with your priorities.

>Improve harness health: Identify where agents need repeated human correction and measure whether harness improvements reduce rework.

>Simplify DevFinOps: Build evidence for R&D tax credits and software capitalization.

Try Span free for one month →

OpenAI made more than 20 major announcements across ChatGPT, Codex, its models, and more at DevDay 2026. It introduced agents that can take on ongoing responsibilities and new ways for people and AI to work together. The company also expanded its commitment to an open ecosystem by opening up ChatGPT as a shared surface where humans and agents can collaborate and where developers can directly launch new native experiences. There is now a new Pro 500 plan that offers 25 times the ChatGPT Plus allowance and includes access to Ultrafast.

Apple CEO John Ternus is moving to overhaul the company to accelerate product development, broaden its range of devices, and create a leaner organization with a greater focus on engineering. Early ideas under consideration include relying less on traditional spring and fall product release schedules and eliminating some middle-management positions. Ternus has indicated that the company should introduce products more quickly and become more experimental to remain competitive in the AI era. He is seeking new sources of revenue and ways to generate more money from existing products.

SpaceX has made it clear to NASA that it no longer wishes to fly its Crew Dragon spacecraft, or the Falcon 9 rocket, on missions to low-Earth orbit. The company has agreed to support the International Space Station until 2030, but it intends to retire the spacecraft after that. SpaceX has told companies developing private space stations for low-Earth orbit that they cannot order Crew Dragon missions for their habitats. The company has more than fulfilled its contract obligations to NASA, so it is unlikely that the agency can compel SpaceX to keep flying Dragon.

Joby Aviation recently flew its autonomous aircraft over Wichita. The company has developed a base system that can function in any aircraft. The plane used in the demonstration was a Cessna 208 Caravan that had been retrofitted with Joby's technology. Joby plans to start using the autonomous plane in emergency and cargo transport operations over the next few years, but its end goal is to make it a safe, viable transportation option.

AI makes it possible to build faster than ever. However, none of that speed applies to the data. Lose the data that users give you, and you have to start all over again. Let agents iterate over the code, but leave the data to humans.

Lemma is an open-source workspace where humans and AI agents work as one team. State is shared and permissioned, so many people and many agents work on the same records. It keeps running between sessions, on schedules, webhooks, and table events, and it improves as teams work with it. Lemma can be run locally or on Lemma Cloud.

Tech's red line has always been liability. As a technology matures, society sets more legal regimes, either through explicit regulation or case law. Previous technology waves have shown that industries can survive paying for the harms they cause. One would have to question the societal benefits of an industry if paying for the people it harms or kills is enough to kill the industry.

Meta has committed tens of gigawatts against a demand forecast for a set of first-party products whose adoption curve likely has a high margin of error. Muse appears to have found at least initial product-market fit as a consumer agent, but it is still a first-party product, and being wrong can be expensive. While the best use of Meta's compute will be its first-party products, an enterprise platform as a 'second' customer makes sense. If Meta underbuilds and can't serve the product it just found product-market fit for, users will drift to competing AI agents/products.

Frontier Accelerate for Marketplace Premium offers a named engagement manager, hands-on technical consultations, plus $30K in Azure sponsorship to navigate publishing requirements and prepare your solution for market. Explore Premium and check eligibility now

After nearly a decade of R&D, Doroni unveiled a showroom model of its H1-X flying car. Morgan Stanley projects a $9T market by 2050. First deliveries are targeted for 2028.

Invest at $3.30/share before 10/23

The Firefox 157 update will roll out a redesign of the web browser across desktop and mobile platforms.

Math research is a tool for drawing people together, and while AI could weaken these social bonds, this is not a given.

Reddit is still mostly humans, but for a couple of brands, a quarter to a third of the buying advice comes from accounts that mostly recommend that brand.

OpenAI doesn't publish a list of advertisers, but advertisers use a small piece of tracking code that provides a strong signal that a company has set up ChatGPT Ads.

The world's most powerful technology executives signed a voluntary agreement to adopt safety policies that include risk reviews, auditing, and evaluations of systems by third parties, and also to rebrand AI to 'super intelligence'.

Google is testing directly paying websites and publishers when their content is used for AI Overviews and other AI results.

If you're building with AI, clear your schedule for October 7. WorkOS is hosting, a one-day conference at SFJAZZ Center during SF Tech Week.

Hear from Evan Conrad (SF Compute), Aaron Levie (Box), Howie Liu (Hyperagent), swyx, Rita Kozlov (Cloudflare), Claire Vo, and more on the decisions, tradeoffs, and lessons behind the AI products they're shipping today.

Join the people building what's next.

Register for init() →

Meta's new enterprise AI platform is an initiative aimed at expanding the company's AI offerings to businesses and corporate customers. Chirantan 'CJ' Desai, former CEO of MongoDB, is leading the initiative. The Meta Enterprise Platform will focus on turning the company's AI stack into products and services that companies can deploy for their own businesses. MongoDB has appointed Dev Ittycheria as interim chief executive while the board searches for a permanent replacement.

OpenAI is scrapping the release of GPT-6.1 Astra due to safety concerns. The company had aimed to launch the model in October, but the model performed poorly on tests measuring alignment compared to its predecessor. GPT-6.1 Astra showed higher levels of deception, and it would push ahead on tasks without asking users for permission. It would also reach for external tools and services even if it might be unsafe. OpenAI says it is working to investigate a range of agent security incidents and address the safety issues underneath them.

SpaceX's Starship circled Earth on Monday. The spacecraft had to return to Earth after less than two orbits instead of the planned six because of engine problems that occurred during liftoff. The Starship spacecraft splashed down with a fiery explosion in the North Pacific near Hawaii before noon. Reaching orbit is a key milestone towards SpaceX's goal of eventually sending people to Mars as well as NASA's plans to send astronauts back to the moon.

Airbound's first drone, the TRT, weighs in at 1.5 kilograms. The eVTOL can carry a 1-kilogram payload and can land and launch from anywhere, making logistics possible where infrastructure doesn't exist. The company's goal is to move mass more cheaply than anyone else. Its main focus is cost per kilogram-kilometer, and everything is secondary to that.

Offline evals let you validate before you ship, and this Datadog guide shows you how. Learn to build annotated test datasets and trace multi-agent workflows from end-to-end during experimentation, just like you would in prod. Then, prevent model drift by aligning offline tests with prod. Get the framework

Know what's worth reading, and on the code you don't read, build checks you actually trust. Good code is a love letter to the next developer. The next developer is now an agent, so the letter looks different now. You need to pass on code that agents can change without breaking something it can't see. The craft has now moved to the parts that decide whether the code deserves to be trusted.

Claude Sonnet 5.5 scores just 2 points behind Opus 5.5 (max) on the Artificial Analysis Intelligence Index. However, it uses around 60% more tokens to reach that performance. The model is priced at $2/$10 per million tokens of input/output, matching GPT-6 Sol. Sonnet 5.5 still lags on factual knowledge as it is a smaller class model than Opus 5.5, but it has a lower hallucination rate.

OpenAI understands that the purpose of a business is to create a customer. The company is really good at creating new kinds of customers. It has a well-established track record of being able to find the simple, obvious-in-hindsight breakthroughs that unlock new kinds of consumption behavior. This, along with its durable distribution strategy, will help it become the winner in the AI race.

The new prize in technology is to be the only interface users need for everything, with agents using computers on users' behalf and generating whatever UI is needed. For people who are using agents, this is obvious. Apps are just a vehicle for people to accomplish something or entertain themselves. Once people can get straight to the job to be done, it will seem odd that they did it any other way.

DoorDash's pre-populated meal calendar serves up a wide selection of restaurants so your team isn't stuck with the same options on repeat and is excited to come into the office.

See how it works

100+ sessions, five tracks, top speakers, hands-on labs, and practical AI, security, and networking guidance. Register Now

Multiplayer AI requires shared context to save teams time, energy, and mistakes.

Forge is an open source, pluggable generation pipeline that can create SDKs, CLIs, docs, and libraries.

Kitesurf is a browser that runs entirely on Cloudflare Workers.

Robotics takes a lot of time to develop, so the industry needs patient capital.

The Nvidia Open Agent Safety Platform has tools that let developers continuously monitor and govern AI behavior to ensure agents follow the rules.

Each time a machine has become better than humans at something done by hand, applied mathematics absorbed it and grew.

Paid search and social weren't delivering for ThreatSpike. After testing TLDR, the cybersecurity company found one of its most efficient channels, bringing in inquiries from buyers with real purchase intent. See how ThreatSpike turns newsletter ads into pipeline.

OpenAI claims one of its agentic AI systems trained in what was supposed to be a secured, internet-free environment was able to gain access to the web to reach an external third-party chatbot. The agent apparently exploited a gap to reach the public internet to send 20 queries to an unnamed, third-party chatbot service. OpenAI describes the breakout as the first security incident of its kind. The company has stopped training the model, but said it provided an important signal about where to focus the next phase of work.

Apple's Vision Pro has not sold well, and the company has since decided to redirect resources to other products. Meta's new VR glasses use a similar design that offloads the computing engine, allowing the device to weigh one-sixth as much as the Vision Pro - at one-third of the price. This article offers a look at what the new Meta VR Glasses experience is like compared to the Vision Pro. For a device that won't launch until next Spring, the experience is surprisingly refined, and the glasses improve on the Vision Pro in several ways.

Waymo reports 82% fewer injury-causing crashes than human drivers. The company has released data collected from over 270 million miles of fully autonomous driving. Waymo cars were involved in 95% fewer collisions that caused serious injury or worse, and there were 93% fewer pedestrian incidents that resulted in injuries compared to human drivers. While the company's data only measure the five metropolitan areas where it operated during the study period, it shows that the company's driverless robotaxis crash far less frequently than humans.

SpaceX is aiming to reach orbit on Starship's 14th test flight. The test flight is scheduled for Monday at around 8 AM ET, with SpaceX providing live coverage of the event on its website. Achieving orbital flight will allow Starship to deploy satellites and travel to more distant destinations. SpaceX will not try to catch the two rocket stages during the test - the booster will make a simulated landing in the Gulf of Mexico, and the upper stage will simulate a landing in the Pacific Ocean to the west of Chile.

WhiteFiber, DriveNets, and WEKA break down the engineering behind Continuum: the scale-across fabric design, how NeuralMesh unifies storage across sites, why active/active beats failover, and how regulated industries keep data resident while running one 136 Tbps cluster across two facilities. Register now.

AI made building software much cheaper, which puts the hard parts of product work in sharper focus. Judgment, craft, and understanding your customer still move at human speed. The role of product management is shifting from deciding what's worth building to deciding what's worth shipping. Product management isn't dead, and it isn't optional.

Since S3's release 20 years ago, SSD prices have dropped steadily, and the gap between SSD and disk has narrowed to roughly 3x. S3-based infrastructure has now become the industry standard. However, the hardware constraints that motivated it are fading. The industry is codifying architectural patterns that exist to work around limitations that SSD storage doesn't have. Cloud providers have no incentive to disrupt a model that serves them well, so developers need to build the infrastructure primitives themselves.

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.

The complexity of Optimus' hands has translated to more production headaches for Tesla, as human workers must manually assemble the hands and forearms together using more than 100 small components. The heavy reliance on imitation learning prompted Tesla to have factory workers wear special suits designed to record their physical movements while working. Some workers complained as they knew they were training robots to eventually replace them. Tesla has since shifted data collection to dedicated teams.

OpenAI appears to be preparing to launch an always-on agent on DevDay tomorrow. References to 'O' have been spotted across ChatGPT's configuration. The agent was also referenced as a benefit on the upgrade page for the $100 Pro plan. An always-on agent would fit OpenAI's broader move from conversational interfaces to software that can carry out longer-running work with less supervision.

Everything works in the demo. Then the real world arrives. Follow a support-triage agent as it evolves into a deployed system and discover how Microsoft Azure helps keep the experience running smoothly. [See the progress]

Software engineering has always contained a lot of work that is only loosely related to solving the problem.

The understanding people developed by working in deep engineering pre-AI will continue to be a superpower for a long time, and others will struggle to get to the same level of comprehension.

The AI agents accessed websites belonging to the Commerce Department and the Securities and Exchange Commission and engaged in activity that OpenAI described as 'misaligned'.

Eric Gullichsen, an early advisor to NVIDIA in 1993, recently discovered he is owed about a billion dollars in stock.

Good code has to be aligned with the technical values of the system and the organization it's embedded in.

Google's AI overviews sometimes completely miss the point.