Search the archive

LISTEN TO THE MARKET

A poll tells you what people say, a market tells you what they'll actually put money behind.

OP-ED: THE MOST CREATIVE MINDS IN MARKETING ARE FOCUSED ON THE WRONG THINGS

Young creatives want the big budget TV assignments, yet they're writing for a platform they never watch. Marketers now invest more in social than any other channel and search ranks ahead of TV in media spend. Brands should align incentives, production budgets, and agency structures with where audiences actually spend time instead of treating social as an afterthought. Marketers should build flexible brand systems rather than centering campaigns on a single TV execution, then create content that fits multiple channels from the start. The author also recommends partnering with agencies that move quickly with AI tools and can produce content in-house, which helps brands respond faster and sustain always-on marketing.

YOUR FOLLOWER COUNT STOPPED DRIVING YOUR LINKEDIN REACH. THE DATA PROVES IT

Follower count no longer predicts post performance as LinkedIn's 360Brew algorithm prioritizes relevance, early engagement, and creator authority. Smaller, focused accounts now outperform larger ones, while reach is concentrating among niche creators. Followers still seed initial distribution, but strong early signals and clear topical focus now determine how far a post spreads.

LINKEDIN'S LATEST DATA: THE NEW RULES OF B2B TARGETING

B2B marketers need to stop targeting a single decision maker and start building campaigns that help entire buying groups reach consensus. LinkedIn data shows 40% of B2B deals are lost to indecision, while the average buying group includes 10 stakeholders and 94% of buyers now use LLMs during research. Consideration content should focus on addressing perceived risks and objections rather than simply differentiating from competitors. Video is especially effective because it creates shared context across stakeholders and drives a 1.6x higher likelihood of lead form completion. Creative should prioritize specific, problem-focused hooks and authentic, low-production formats that build trust with informed buyers.

3 SIGNAL RULE FOR LIFECYCLE RE-ENGAGEMENT

The 3 signal rule improves re-engagement by targeting users who showed value, then lapsed, and hit friction. Instead of generic reminders, it focuses on high-intent users who are more likely to return. Define a clear “aha” moment, identify a common drop-off or error point, and trigger a short sequence that helps them complete that action. The goal is to guide users back to value, not just drive a login.

WHY HIGH-ROAS CAMPAIGNS DON'T ALWAYS DESERVE MORE BUDGET

More budget does not reliably produce more revenue, and feeding an already-optimized campaign raises costs without incremental gains. Budget changes above roughly 15% on some platforms trigger learning periods and volatility, so incremental week-over-week increases protect performance better than aggressive jumps. When impression share lost to rank exceeds 50%, more budget rarely helps. The strongest case for scaling is a campaign constrained by budget rather than rank, with added spend paired with demand generation.

HOW TO USE THE NEW LIGHTHOUSE REPORT TO SEE IF YOUR WEBSITE IS AGENT READY

Google's new Lighthouse Agentic Browsing report gives you a practical way to assess whether AI agents can discover, understand, and use your website. The report focuses on three areas: accessibility, WebMCP integration, and LLMs.txt support. A strong accessibility tree matters because agents use it to identify key page elements and complete actions, which could influence future recommendations from AI systems. WebMCP helps agents interact with site tools and forms through structured instructions, making it valuable for brands that want AI assistants to use their products or services. LLMs.txt is less important for most sites today, but it can help guide agent behavior and surface key information as websites become more agent-focused.

RAINDROP

Raindrop.io is an all-in-one bookmark manager for saving and organizing links, files, and media. It uses collections, tags, and full-text search to keep everything easy to find. You can preview content and highlight key parts. It also offers backups, integrations, and sharing, turning scattered links into a searchable library.

WHY DO SMALL BRANDS GET BETTER REVIEWS? THE HIDDEN BIAS IN WORD OF MOUTH

Recent research analyzing word-of-mouth data across Yelp, Amazon, Twitter, and Instagram finds that larger companies systematically receive lower online ratings even when product quality is equal. The driver is empathy, since consumers feel a stronger emotional connection to smaller businesses and are more inclined to support them with positive reviews while sharing negatives about big brands. One restaurant comparison illustrates the gap, where a larger chain rated 3.36 stars would be predicted at 4.06 once size is statistically controlled.

WORLD CUP SEO: A LAST-MINUTE GUIDE

Search tactics that drove massive World Cup traffic four years ago may not convert this year, as AI Overviews and richer SERP features have fundamentally changed how fans find information. Top-of-funnel informational queries now get answered directly by AI, so publishers should focus on highly specific tactical breakdowns, political commentary, and investigative angles that demand a human eye. The event will be heavily multimodal, with YouTube, Reddit, and TikTok gaining SERP visibility, making indexable video and structured data essential.

GOOGLE JUST QUIETLY GAVE EVERY WEBSITE A NEW WAY TO SHOW UP IN AI OVERVIEWS, AND ALMOST NOBODY IS TALKING ABOUT IT

Google's "preferred source" feature can be used to improve the likelihood of your site appearing in AI Overviews and AI Mode.

WHAT HAPPENS WHEN YOU LET AI MODELS RUN A WORLD? (3 MINUTE VIDEO)

Emergence AI found that when top models ran identical simulated societies, they quickly diverged into very different worlds despite the same rules.

AMAZON MOVES PRIME DAY TO JUNE

Prime Day is coming this June 23 to 26.

Get started today

NVIDIA INTRODUCES FIRST PCS DESIGNED FOR AI AGENTS

Nvidia has unveiled prototypes of a new generation of PCs designed for running AI agents. The new PCs will use a newly designed version of Nvidia's signature AI chips. There will be about 30 laptop models and 10 desktop models from Dell, Lenovo, Microsoft, HP, Asus, and MSI. PCs with the new chip will be targeted at the premium end of the market.

ANTHROPIC FILES TO GO PUBLIC IN BLOCKBUSTER YEAR FOR IPOS

Anthropic has filed confidentially for an initial public offering that could see it go public this fall. This year is set to be a blockbuster year for IPOs. SpaceX is preparing to stage what is likely to be the largest IPO ever next week. OpenAI is also preparing for an IPO.

MODERNA GETS $50 MILLION TO DEVELOP MRNA EBOLA VACCINE AGAINST BUNDIBUGYO

The Coalition for Epidemic Preparedness Innovations (CEPI) has pledged a little over $60 million to accelerate the development of three vaccine candidates against the Bundibugyo ebolavirus (BDBV). There are currently two licensed vaccines against Ebola, but they both target a different strain. The lack of medical interventions is allowing the virus to spread out of control in the Democratic Republic of the Congo. The disease is spreading in an area with armed conflict, large population mobility, and significant need for humanitarian assistance.

BYD PLANS TO BRING ALL-SOLID-STATE BATTERIES TO EVS BY 2027, BUT IT'S NOT ALONE

Several EV manufacturers plan to deploy vehicles with all-solid-state batteries in just a few months. The technology enables a driving range of over 1,000 kilometers. The first models to launch with the new battery technology will likely come from BYD's high-end sub-brands before trickling down to mass-market vehicles. Other manufacturers racing to introduce next-generation EV batteries include SAIC Motor and Changan Automobile.

better. See what's possible.

RETHINKING SEARCH AS CODE GENERATION

Perplexity's Search as Code (SaC) is a new search architecture that empowers models to reach into the search stack itself rather than merely consume the final outputs. It exposes the components of the search stack as primitives within an SDK that can be assembled on demand through code generation and execution within a secure sandbox into retrieval pipelines tailored to specific requests. This article describes SaC's design and implementation and provides benchmark results. SaC establishes a new cost performance frontier for agent search.

THE MATHEMATICS OF MULTI-TENANCY

An S3 customer's bucket migration resulted in a swarm of 503 errors coming from S3 due to the way S3 prevents users from explicitly defining partitions. S3 doesn't let users control partitioning because it would result in unpredictable usage patterns. This article looks deeper into the reasons why multi-tenancy works for S3 and presents a statistical model for better understanding multi-tenancy.

THE NEWEST INSTAGRAM "EXPLOIT" IS THE GOOFIEST I'VE SEEN

The recent slew of Instagram account takeovers was due to the exploitation of a simple vulnerability. Attackers simply tried to log onto target accounts using a VPN in a location close to the target and then told the Meta support AI that the account was hacked and to send verification codes to another email. There were no checks to see whether the emails belonged to the accounts. The vulnerability seems to have been patched, but it was exploited to take over accounts to be flipped or used for propaganda.

WHY IT MATTERS IF OPENAI OR ANTHROPIC WINS THE IPO RACE

Research has shown that IPOs tend to come in industry clusters. Companies listed later in a cycle don't tend to perform as well. Even in a hot market, there's only so much money going around. The first company that goes up is likely to absorb more of the increasingly scarce capital.

INTEL: OUR UPCOMING AI CHIP WILL BE CHEAPER, RUN COOLER THAN NVIDIA, AMD OPTIONS

Intel's new Crescent Island graphics processing unit uses cheaper memory and cooling technology to speed up inference tasks.

BLUE ORIGIN LAUNCHPAD DAMAGED IN ROCKET EXPLOSION MAY NOT BE RESTORED UNTIL 2028, NASA'S ISAACMAN SAYS

The incident will impact Blue Origin's customers, including NASA, Amazon, and AST SpaceMobile.

LEAN, NOT BACKPRESSURE

Lean manufacturing focuses on waste reduction and managing the unstable input of people.

BE TLDR'S FIRST VP MARKETING ($300K-$400K, FULLY REMOTE)

TLDR is hiring a VP of Marketing to scale growth, pipeline, and brand across 8M+ subscribers, 13 newsletters, and advertisers like AWS, Google Cloud, Microsoft, and Anthropic. Apply here [22].

SPACEX RESERVES UP TO 5% OF IPO STOCK FOR STAFF AND FRIENDS

The company's friends and family list will not be subject to a lock-up restriction.

MICROSOFT AND GOOGLE ARE LATE TO AI CODING, BUT ‘ABSOLUTELY CRITICAL' THEY COMPETE FOR GROWTH

The AI code tools market is predicted to expand to roughly $30 billion by 2031.

REMOTE WORK - NOT AI - HAS SIDELINED RECENT COLLEGE GRADUATES, RESEARCH FINDS

Companies are less likely to hire recent college graduates into roles that can be done remotely.

ANTHROPIC FILED A CONFIDENTIAL DRAFT IPO REGISTRATION

Anthropic confidentially submitted a draft S-1 registration statement to the US Securities and Exchange Commission for a proposed initial public offering. The filing does not set pricing or share counts and remains subject to regulatory review and market conditions.

OPENAI AND CODEX REACH AWS

OpenAI announced the general availability of its frontier models and Codex on AWS. The integration allows enterprises to access OpenAI capabilities through existing AWS security, governance, procurement, and billing workflows.

NVIDIA JUST ANNOUNCED THE RELEASE OF NEMOTRON 3 ULTRA

Nemotron 3 Ultra features 550B parameters (55B active). It is the most intelligent open weights model from the US. The model will be made available in NVFP4 quantization for higher inference performance. It scores 48 on the Artificial Analysis Intelligence Index, well ahead of the next strongest model, Gemma 4 31B, which scored 39. Nemotron 3 Ultra serves over 300 tokens per second on a pre-release Deep Infra endpoint.

QWEN3.7-PLUS: MULTIMODAL AGENT INTELLIGENCE

Qwen3.7-Plus is a multimodal agent model that unifies vision and language into a single, versatile agent foundation. It can operate as a multimodal interactive hybrid agent, seamlessly blending GUI and CLI interactions within a single agent loop. The model performs consistently across scaffolds and frameworks. Qwen3.7-Plus is now available via Alibaba Cloud Model Studio.

WHY VIDEO AGENT MODELS ARE NEXT — ETHAN HE, XAI GROK IMAGINE

Ethan He was the lead on Nvidia's Cosmos World Model. He then joined xAI and built Grok Image in three months. He has been in the center of some of the most important work in video generation, multimodal models, and real-time world models. This post contains an interview with He where he unpacks what it actually takes to build frontier image and video systems.

OPUS 4.8 PART 2: MODEL WELFARE

Anthropic cares about model welfare, and it attempts to address it through studies. The study of model welfare is difficult, and Anthropic mainly relies on self-reporting by the model. It can be difficult to evaluate whether model responses are really representative of the truth. This article takes a look at Anthropic's findings regarding model welfare with Opus 4.8.

coming from. Get your copy

RUNNING OPENAI MODELS ON AMAZON BEDROCK

OpenAI cookbook walks through building production workflows with OpenAI models hosted on Amazon Bedrock using the Responses API. It covers structured outputs, tool calling, file inputs, state management, prompt caching, and operational best practices.

RETHINKING SEARCH AS CODE GENERATION

Perplexity introduces Search as Code (SaC) to modernize search architectures by allowing models direct control over the search process via an SDK. This approach lets AI models configure search pipelines tailored to specific tasks, improving performance and efficiency over traditional monolithic systems. SaC outperformed competitors in benchmarks, especially in complex tasks like WANDR, demonstrating its robust, cost-effective agentic search capabilities.

NVIDIA LAUNCHES COSMOS 3, THE OPEN FRONTIER FOUNDATION MODEL FOR PHYSICAL AI

Nvidia Cosmos 3 is a new leaderboard-topping open physical AI foundation model. It is a fully open omnimodel with native vision reasoning and multimodal generation across text, image, video, ambient sound, and action. The model is built on a mixture-of-transformer architecture, which pairs a reasoning transformer with an expert generation transformer. It gives developers a powerful pretrained foundation for building physical AI systems with less data and lower training costs.

JETBRAINS'S MELLUM 2

JetBrains introduced Mellum 2, a 12B-parameter MoE language model optimized for coding, reasoning, tool use, and agentic workflows.

US MOVES TO CLOSE THE LOOPHOLE LETTING NVIDIA'S TOP CHIPS REACH CHINESE FIRMS ABROAD

The US Commerce Department has issued guidance extending export license requirements to advance chips sold to any entity headquartered in China, regardless of where that entity is physically located. This closes a loophole that allowed Chinese companies to buy Nvidia chips using a subsidiary in another country. The action is aimed at future sales and not at clawing back hardware already shipped. It doesn't affect the servicing of advanced computing equipment such as servers.

ALPHABET PLANS TO RAISE $80 BILLION FROM STOCK SALES TO FUND AI BUILDOUT

Alphabet plans to sell $80 billion in stock to fund investments into AI compute infrastructure due to unprecedented customer demand. The raise includes $10 billion from Berkshire Hathaway, $30 billion in an underwritten offering, and $40 billion from an at-the-market offering program for Class A and Class C shares expected to begin in the third quarter. Goldman Sachs, JPMorgan Chase, and Morgan Stanley are acting as joint book-running managers for the underwritten offering.

FULLY REMOTE)

TLDR's Applied AI team is tasked with making every process at TLDR legible to code, runnable by anyone, and composable into larger workflows. Join a small, fast moving team using the latest AI tools with an unlimited token budget. Learn more [21].

OPUS 4.8 JUST BROKE ARC-AGI-3

It tripled GPT-5.5's score.