OpenAI's custom chip 🖥️, Tesla virtual power plants⚡, coding token costs 💰

OpenAI and Broadcom's Jalapeño chip is designed for inference, the compute-intensive process of serving AI models to users ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌  ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 

Sign Up |Advertise|View Online TLDR Together With TLDR 2026-06-25 Black Friday hit at 9:47am. NiCE AI Agents handled the next 4.2 million conversations (Sponsor) Do you love listening to hold music on customer support calls? Your customers probably don't... With NiCE AI Agents, they never need to hear it again, turning 40-minute waits into 40-second resolutions. Proven at enterprise scale, these AI Agents connect your organization's knowledge to customer needs, automating the full customer journey from intent to resolution. Most AI was built to demo. NiCE AI Agents are built for deployment, consistently enhancing every interaction. Anyone can build an AI Agent. NiCE runs them at scale. Get started 📱 Big Tech & Startups OpenAI unveils first chip as part of Broadcom deal in effort to ‘build the full stack' (6 minute read) OpenAI and Broadcom's Jalapeño chip is designed for inference, the compute-intensive process of serving AI models to users. It took nine months for the companies to design the chips end-to-end. The use of OpenAI's AI models helped accelerate the process. Jalapeño is a major step in OpenAI's plan to build the full-stack behind its models and products. Tesla, Sunrun team up on 16 GW virtual power plant for data centers (6 minute read) Sunrun, Tesla, and Renew Home have announced an agreement to aggregate more than 16 gigawatts of home batteries and other devices into the largest distributed power plant in the US. The project aims to address the surging electricity demand from data centers. The companies already have more than 300 megawatts of capacity ready for immediate deployment in Virginia. They expect capacity to grow to at least 500 megawatts by 2030 as more home batteries and thermostats come online. 🚀 Science & Futuristic Technology A New $500 Million Fund Is Trying to Eliminate the Common Cold (5 minute read) Intercept is a project aimed at making respiratory infections like the common cold and flu a thing of the past. Launched on June 24, it will support research into the adoption of technologies like air filtration to remove pathogens from the air directly. The project has received $500 million in investments from a variety of sources, including Stripe and Anthropic. The funds will be used to back the development of treatments that could protect people from many viruses at once. The Roomba Guy's Second Act: A Robot You'll Want to Snuggle (9 minute read) Colin Angle, who led iRobot for nearly three decades, has a new startup called Familiar Machines & Magic that is selling a robot designed to help people monitor their loved ones. The robot, called the Familiar, is an expressive, furry creature that can react to people's actions and feelings in an emotionally intelligent way. It is still in an early prototype stage, so there's no information about price or availability, but videos of the current iteration of the product are available in the article. The Familiar will use local small models to process requests and won't send any data to the cloud. 💻 Programming, Design & Data Science How Baseten built the world's fastest API for GLM 5.2 (Sponsor) For coding and agentic workloads, GLM 5.2 delivers competitive intelligence to Opus 4.8 at 1/5th the cost. Baseten's GLM 5.2 API delivers 280+ TPS, <0.8 TTFT with 99.9% uptime. This is a great model to achieve higher uptime and lower costs vs. closed weights. Try GLM 5.2 What I'm Finding About LLM Code Style and Token Costs (16 minute read) Large language models will code from scratch unless you explicitly tell them about existing functionalities. They don't know to use existing code unless you tell them to. Letting them know what already exists can result in significant token savings. It can also help remove entire categories of security and reliability issues. The Case for Language-Native Software (12 minute read) Natural language is becoming an interface. Software is beginning to understand instructions in human language. Language-native software will accept intent in the language users already speak, translate it into structured operations, and execute them deterministically. This makes language the interface itself. 🎁 Miscellaneous TLDR is hiring a Senior PMM ($180k-$225k base + $40-50k annual target bonus, Fully Remote) We're hiring a senior PMM to own product marketing at TLDR. You'll define our positioning, build out sales enablement, and lead every launch. Learn more. The CEO of AWS on why Amazon is hiring 11,000 interns and junior employees (22 minute read) Amazon is increasingly in the business of selling software that does white-collar work. However, Matt Garman, the CEO of AWS, says he isn't worried about AI destroying a huge number of jobs. He believes that jobs might change, but they're not going away. This article features highlights from an interview with Garman where he discusses how technology changes work. Anthropic's White House Negotiations Are Reportedly On Track After ‘Weirdo' Dario Amodei Was Replaced (2 minute read) Dario Amodei was reportedly weird and hard to deal with during the negotiations between the White House and Anthropic. Talks are now on track now that his role in the discussions has been replaced by Tom Brown, another Anthropic co-founder. Brown's manner is warmer and more traditionally personable. He is working on the talks alongside Sarah Heck, Anthropic's Head of Public Policy, who is quite disciplined and cautious but comes across as enthusiastic and on-message. ⚡ Quick Links The notetaker bot has met its match (Sponsor) Granola's AI notepad listens in without joining the call and drafts next steps from your POV. Get 1 month free with code TLDR1MO. Stop Programming in Markdown (7 minute read) People are using LLMs inappropriately in situations where regular code would fare much better. Qualcomm lands Meta as first named customer for its Dragonfly data centre chips (5 minute read) The Dragonfly C1000 is a general-purpose server processor expected to be available in 2028. Amazon Seller Reveals Rare Glimpse of Shadow Bribery Market (11 minute read) Middlemen on messaging apps like Telegram, WeChat, and WhatsApp are offering access to people inside Amazon who can get things done for a price. Reach 7 million engaged tech professionals (Sponsor) 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. Writing Loops, Not Prompts, Explained (15 minute read) Automate the parts of prompting that you keep repeating, and keep judgment close to the parts that matter. Stop Building Chatbots. Build Agents That Open PRs (14 minute read) PRs are an artifact that can be reviewed or rejected by a human. I wrote a 70x faster SQL parser while barely looking at the code (8 minute read) Lessons learned when rewriting PostHog's SQL parser using multiple Claude Code sessions running in parallel. Love TLDR? Tell your friends and get rewards! Share your referral link below with friends to get free TLDR swag! https://refer.tldr.tech/7c97ce4b/ Track your referrals here. Want to advertise in TLDR? 📰 If your company is interested in reaching an audience of tech executives, decision-makers and engineers, you may want to advertise with us. Want to work at TLDR? 💼 Apply here, create your own role or send a friend's resume to jobs@tldr.tech and get $1k if we hire them! TLDR is one of Inc.'s Best Bootstrapped businesses of 2025. If you have any comments or feedback, just respond to this email! Thanks for reading, Dan Ni & Stephen Flanders Manage your subscriptions to our other newsletters on tech, startups, and programming. Or if TLDR isn't for you, please unsubscribe.