Search the archive

Learn about our methodology [5] → · Explore R&D roles at Garner

→

SPACEX PLANS TO LAUNCH STARLINK MOBILE SERVICE IN THE US

SpaceX plans to launch a new Starlink mobile service for customers in the US. The company is considering building its own terrestrial mobile network in the country. The new retail offering would put SpaceX in direct competition with Verizon, AT&T, and T-Mobile. Some analysts say that the idea may be a gamble to extract better deals from the company's telecom partners, as it would cost billions of dollars to roll out the required infrastructure.

ZUCKERBERG URGES META TO EXPLORE WORKING WITH POLYMARKET AND KALSHI

Meta is building an app that allows people to make bets on practically anything, similar to Polymarket and Kalshi. Arena, which is now a top priority within Meta, will use points instead of real money and is aimed at 18- to 34-year-olds. The company is now actively trying to form partnerships with Polymarket and Kalshi, but it is unclear what kind of arrangement Meta is seeking. If released, Arena will be a stand-alone app, but Meta eventually plans to integrate parts of it into Facebook and Messenger.

WHY THE WEST STOPPED MAKING LAND

Around 8% of the land in the major coastal cities in the US was underwater in the 1890s. Some cities, like Boston and Charleston, have doubled in size by reclaiming land. The demand for land is now higher than ever, but the country has virtually ceased to reclaim any at all. Regulations introduced in the 1970s made it enormously expensive to reshape the landscape. Lifting these legal barriers could mean hundreds of thousands of new homes near the centers of the country's most valuable cities.

APPLE AND AUDI ALUMNI HAVE MADE A LUXE EV BASED ON THE MOON BUGGY

Amble, a European electric car company, has released a $25,000 electric buggy designed for places where a normal car feels out of place. The Amble One has a range of more than 60 miles with a top speed capped at 40 miles per hour. It takes five hours to charge the vehicle from any standard home socket. The vehicle weighs under 450 kilograms, which allows it to drive on public roads in Europe without being treated as a car.

USING LOCAL CODING AGENTS

A local setup is a viable alternative to using proprietary services for many coding workflows. Running agents locally makes them transparent, inspectable, and free to run apart from hardware and electricity costs. This tutorial walks readers through how to set up a production-ready coding agent with a fully local stack. The agent uses a locally served LLM with a coding harness that allows it to read files, make edits, run commands, and verify changes.

FINTECH ENGINEERING HANDBOOK

This handbook describes the most important patterns used in software engineering when money is the primary focus of the system. It aims to provide a comprehensive understanding for people entering fintech, those already in it, and those outside the industry. The guide follows three principles: no invented data, no lost data, and no trust. Money can't be created out of nowhere, everything that happens to money has to be tracked and persisted, and everything must be verified.

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 [14].

INTEL'S CHIP BUSINESS SHOWS SIGNS OF LIFE AFTER YEARS OF STRUGGLE

Intel was flailing until the US government intervened and took a 10% stake in the company last summer. The company is now showing signs of a turnaround, with its value having more than tripled. Intel has since added big customers like Nvidia and Apple. The AI boom has reshaped the chip market and benefited Intel greatly.

TRUMP ADMINISTRATION ROLLS BACK PART OF ANTHROPIC MODEL BAN

Anthropic is now allowed to serve its Mythos 5 model to trusted companies and government partners. Fable 5 remains restricted and restrictions on Mythos 5 still apply to entities that aren't trusted partners. The US government is racing to implement an executive order recently signed that gives federal cybersecurity officials a say in AI model evaluation. The industry is in limbo and will have to react to case-by-case decisions by the administration until the order is implemented and standards are in place.

AI IS MAKING SILICON VALLEY PRODUCTIVE, ANXIOUS, AND AFRAID TO LOG OFF

People using AI agents are working longer hours than before due to anxieties about how AI might advance without them if they log off.

GOOGLE IS RATIONING GEMINI ACCESS TO META BECAUSE IT CANNOT PROVIDE ENOUGH COMPUTE

Meta has told staff to use AI tokens more efficiently and shift to using the company's own Muse Spark model.

AGENTICS/TECH THINGS: TOKENMAXXING IS DEAD, LONG LIVE TOKENMAXXING

Tokenmaxxing the top labs will never stand up to any amount of CFO scrutiny, but using open models may shift the cost-benefit equation.

AN “INFOVORE” SHARES HIS CHATS

Tyler Cowen is a polymath who uses ChatGPT to pursue his endless curiosity.

DEFINING TASTE

Taste is the ability to consistently make high-quality qualitative judgments where no objective metric exists.

WHAT HAPPENED AFTER 2,000 PEOPLE TRIED TO HACK MY AI ASSISTANT

Prompt injections are still a real security problem, but defenses are getting better.

Read the story behind the two-day modernization

GPT-5.6 SOL, TERRA, AND LUNA

OpenAI introduced GPT-5.6 Preview, a family of models named Sol, Terra, and Luna, with Sol positioned as the flagship model. The system card describes stronger cyber and bio safety testing, new safeguards, and a limited preview before broader availability.

MUSK SAYS GROK 4.5 ENTERED PRIVATE BETA

Elon Musk said Grok 4.5 was in private beta at SpaceX and Tesla. The model is based on a 1.5T V9 foundation model with Cursor data added during supplemental training. Early evaluations were near or above Opus, with reinforcement learning still improving the model.

GOOGLE LIMITING META'S GEMINI USE

Google reportedly limited Meta's access to Gemini capacity after Meta requested more compute than Google could provide. The shortfall was said to have delayed some internal Meta AI projects and pushed staff to use AI tokens more efficiently.

LEAN SOFTWARE SCALING LAWS

Codebases and programming languages will eventually become easier for AI models to understand, fix, and write. The Lean programming language has a worse baseline constant and total loss on existing code bases compared to other languages, but better scaling components. This implies that implementations in Lean could eventually win and deliver large benefits in program correctness at global scale. This may justify large-scale investments in rewriting existing codebases in Lean or paying for new Lean code.

MEMORY PRICES REPORT FROM STANFORD

Stanford published an interactive report on historic and current memory and storage prices. The dataset follows the spirit of John C. McCallum's classic memory-price work, with downloadable raw data and chart export tools.

THE NEXT PARADIGM

AI labs are betting that scaling reinforcement learning from verifiable rewards (RLVR) across millions of diverse tasks will achieve artificial general intelligence, but this paradigm hits a wall in domains that lack deterministic simulators. True continual learning requires moving past temporary in-context memory and shifting back into the model's weights themselves.

MONEYBALL FOR PHYSICAL AI

Data engineering pipelines should deprecate cumulative operational hours as a primary metric. Engineering efficiency and model scaling should be evaluated using quantifiable parameters. An optimal capital allocation strategy balances data types against their specific utility metrics. Capital efficiency scales by accurately pricing data novelty.

Download the report

ACCELERATING GEMINI NANO MODELS ON PIXEL WITH FROZEN MULTI-TOKEN PREDICTION

Models like Gemini Nano and Gemma make it possible to have powerful large language models right in your pocket. Delivering these models on mobile was a significant challenge. Google built a new architecture that retrofits Multi-Token Prediction onto existing 'frozen' Gemini Nano v3 models to overcome the bottleneck. The new architectural components were designed to maximize efficiency gains specifically for mobile environments. This article shows how Google's research team tackled the unique, extreme constraints of edge computing.

QWEN IMAGE AGENT

Qwen-Image-Agent improved text-to-image generation by planning, reasoning, searching, using memory, and incorporating feedback to fill in missing user context. The work also introduced IA-Bench to test agentic image generation across planning, reasoning, search, and memory.

REWARD MODELS CAN BE TOO SENSITIVE

Meta studied how reward models can overreact to equally good responses, leading reinforcement learning toward reward hacking. The paper proposes measuring both discriminative ability and specificity, then using Monte Carlo dropout to cluster rewards into safer discrete signals.

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 [20].

CLAUDE CODE TURNED EVERY ENGINEER INTO THREE. NOW COMPANIES NEED MORE PRODUCT THINKERS

AI coding agents have dramatically increased engineering output, shifting the bottleneck from writing code to deciding what to build. As software development becomes more automated, engineers who combine strong technical fundamentals with product judgment, customer insight, and code review skills are becoming increasingly valuable.

AGENTS AS WEBS OF BELIEFS

A proposed framework models intelligent agents as interconnected webs of beliefs, where beliefs, goals, and actions emerge from the same underlying structure instead of being treated separately. The approach argues that reasoning, planning, and decision-making arise from maintaining locally consistent belief networks, offering an alternative foundation for building more capable AI agents.

see. No surveys. No guesswork. Take a tour.

APPLE VISION PRO EXEC REPORTEDLY LEAVES FOR OPENAI

Apple Vision Pro executive Paul Meade reportedly left Apple to join OpenAI's hardware team.

WHY ONE OF TECH'S BIGGEST GAMBLERS IS BETTING AGAINST ELON MUSK'S AI VISION

Masayoshi Son thinks the math doesn't support space-based data centers.

GOOGLE TESTS NOTEBOOK COLLECTIONS FOR NOTEBOOKLM

ANTHROPIC ECONOMIC INDEX JUNE 2026 REPORT

Anthropic's June 2026 Economic Index report reveals that AI computational costs strongly correlate with the economic value of tasks, with higher-wage occupations consuming up to 2.5 times more tokens.

READ HOW HEADSPACE DID IT

WHILE JONY IVE IS DESIGNING FERRARIS, ANOTHER APPLE DESIGNER IS THINKING DIFFERENT ABOUT THE GOLF CART

Former Apple designer Julian Hoenig has publicly launched Amble, an electric buggy company whose debut vehicle – the Amble One – is a minimalist, street-legal EV with a 40 mph top speed, 60-mile range, and a starting price of around $25,000. Inspired by retro vehicles and NASA's 1971 lunar rover, the design was born from co-founder José António Uva's frustration with ugly, uncomfortable golf carts in the hospitality industry. All 2027 delivery slots are already reserved for hospitality clients, with individual orders now open in Europe and the US, and deliveries for those expected in 2028.

ADOBE IS BUYING TOPAZ LABS, THE AI VIDEO ENHANCER

FIGMA'S DESIGN AGENT, NOW WITH CUSTOM TOOLS AND GREATER CONTEXT

Figma has expanded its AI-powered design agent in open beta, enabling designers to create custom plugins, WebGPU shaders, and reusable workflows with natural language while giving the agent more context about their projects and teams. The agent can now reference attached files and other Figma documents, search the web, connect to external tools like GitHub, Slack, Atlassian, Notion, and Hex, and save frequently used prompts as shareable Skills. Together, these updates transform the design agent from a simple prompting tool into a context-aware collaborator that helps teams automate workflows, preserve design knowledge, and produce designs that better reflect their brand and way of working.

THE LAYERS OF AI EXPERIENCE

Generative AI has made product systems probabilistic and multi-dimensional, rendering the traditional deterministic design model insufficient. A new framework of six interdependent layers (user interface, context, harness, model, governance, and emergence) shows that design influence now extends well below the surface interface. Effective AI designers must be fluent across all these layers, from shaping model behavior and context engineering to accounting for unpredictable emergent outputs they can't fully control.

WHAT DOES FIGMA DO NEXT?

Figma revolutionized product design by turning the canvas into a multiplayer collaboration space, displacing incumbents like Sketch almost overnight. As AI collapses the gap between visual intent and working code. However, the canvas risks becoming peripheral, with coordination infrastructure, design systems, and live artifacts mattering more than static design files. Figma now faces a strategic fork: keep pulling everything back into its canvas, or accept a humbler role as one collaborative lens into a system where code and runtime are the real source of truth.

TOP ILLUSTRATION AGENCIES SHARE THEIR TIPS ON NEGOTIATING CONTRACTS

Licensing agreements allow illustrators to retain copyright, earn additional fees from future uses, and protect the long-term value of their work, while work-for-hire contracts require significantly higher compensation because they transfer ownership. Creators should carefully review contracts for clauses covering copyright transfer, derivative works, AI training rights, broad exclusivity, and purchase order terms, as these can substantially affect future earnings and control over their work. Understanding licensing and negotiating fair terms can generate recurring income over the course of a career and prevent costly mistakes.

SHADER LAB

A powerful toolkit to create, stack, and animate shaders. Add your first keyframe from the properties panel.

AI THUMBNAIL MAKER FOR YOUTUBE

Generate studio-quality YouTube and Shorts thumbnails that stop scrolling and drive clicks in seconds.

AI-GENERATED VIDEO CREATION PLATFORM

Generate polished, multilingual avatar videos from scripts, briefs, decks, or documents.