Search the archive

SPECFLUENCING: HOW ‘UNOFFICIAL' BRAND MENTIONS ARE CHANGING INFLUENCER CREDIBILITY

Specfluencing is growing fast as creators tag brands without paid deals to appear established and attract partnerships. Between June and September 2025, more than 14,000 influencers posted nearly 77,000 brand tagged posts and generated about 1.2 billion views, and most posts were unpaid. Experts say this tactic works only when content reflects real product use and delivers results, since audiences can detect staged credibility. Brands can use specfluencing to spot new creators and gain unpaid visibility, but it can also create noise and weaken trust if performance is poor. Several brands report that organic creator posts and viral moments can drive stronger trust, engagement, and sales than traditional paid endorsements.

AMAZON TO LAUNCH A MARKETPLACE FOR PUBLISHERS TO SELL CONTENT TO AI COMPANIES

The proposed marketplace would formalize AI-content relationships, shifting from unlicensed web scraping to negotiated data access. Publishers would gain compensation and clearer usage controls, while AI firms would reduce legal risk and secure reliable, high-quality data pipelines. The initiative places Amazon in competition with similar licensing efforts from Microsoft, signaling that structured data marketplaces may become a core layer of the AI economy.

POSITIONING VS. MESSAGING

In PMM interviews, many candidates give answers that lack precision when asked about positioning and messaging. The two are distinct, and your response should reflect that. Positioning is internal and defines the target audience, category, and differentiation. Messaging is external and translates that foundation into benefits and proof points. Tailor your answer to the focus of the question, walk through inputs, process, validation, and activation as relevant.

HOW TO USE EXCLAMATION POINTS IN EMAILS

Using exclamation points in emails makes the sender seem friendlier and more likable, but it reduces how powerful and analytical the sender appears in serious or task focused messages. Across five experiments with over 2,000 participants, messages with exclamation points were rated 13.5% more positive and 21.3% warmer for meeting requests. The same messages were seen as 9.7% less powerful and 12% less analytical in assignment or data-focused contexts. The effect works for both men and women, though women feel 9.7% more pressure to use them and men feel 17.1% more pressure to avoid them. Exclamation points work best in friendly and relationship-building emails and should be avoided in formal, analytical, or deadline-driven messages.

👉 REGISTER NOW

HOW TO TELL DEEP RESEARCH WHICH WEBSITES TO USE FOR ITS RESEARCH

The ChatGPT Deep Research update allows specifying which websites to use for research. It can be activated by clicking the “+” under the prompt box or using the left panel Deep Research button. Once active, the “Sites” dropdown lets users limit searches to specific sites or prioritize them while still scanning the full web.

END-TO-END EVENT PLANNER USING CLAUDE (4 MINUTE VIDEO)

A video walkthrough of an event planning app (V1) built by Tofu's co-founder using Claude Cowork. The app connects Claude, HubSpot, and Clay to automate and streamline many of the manual tasks involved in hosting events.

TRUSTWORTHY A/B PATTERNS: ROUNDED/SQUARE CORNERS PATTERN

Large-scale replications show that rounding button corners does not actually produce a large lift in clicks. The original study reported a 55% increase in clickthrough rate from rounded buttons and had a p value of 0.037, but later researchers judged the test to be underpowered. Three replication tests found effects near zero with 95% confidence intervals. A fourth pre-registered test at METRO Russia ran for 16 weeks with over 7.4 million users, and it found results below the minimum detectable effects. The combined evidence suggests the original 55% lift was likely an exaggerated result caused by low statistical power.

WHY YOUTUBE SHOULD BRING ORIGINALS BACK

YouTube has surpassed Netflix in revenue and viewing time but still lacks big TV advertisers. Reviving original content could provide premium, predictable programming to attract advertisers and retain creators as competitors secure exclusivity deals. With serialized content from creators and more viewers watching on TV screens, the platform is well positioned to explore Originals again through creator partnerships or licensing.

DIGITAL PR STRATEGY: 2 HOURS OF WORK FOR 10+ LINKS

The author spotted an emerging trend, validated the trend on TikTok, wrote a simple pitch, and then built a hyper-relevant media list.

LINKEDIN IMPRESSIONS (JANUARY 2026 BENCHMARKS)

With 5-10K followers, 1-2K impressions places you in the top 25%.

THE NEW FABIO IS CLAUDE

Coral Hart, a prolific South African romance novelist, uses Claude to produce novels in months rather than years.

See what's next

ELON MUSK WANTS TO BUILD AN AI SATELLITE FACTORY ON THE MOON

Elon Musk has told employees at xAI that the company needs a factory on the moon to build AI satellites. These satellites will be launched into space using a massive catapult called a 'mass driver'. Musk plans to build a self-sustaining city on the moon as a steppingstone to Mars and beyond. SpaceX is preparing for an initial public offering that could come as early as June.

MUSK'S XAI LOSES SECOND CO-FOUNDER IN TWO DAYS

Jimmy Ba and Tony Wu, two of xAI's co-founders, have left the company. The departure comes as xAI faces regulatory probes in multiple jurisdictions across Europe, Asia, and the US regarding the company's Grok AI chatbot allowing the mass-creation and syndication of non-consensual, explicit images. These images were based on photos of real people, including children. xAI merged with SpaceX earlier this month.

THE ISOMORPHIC LABS DRUG DESIGN ENGINE UNLOCKS A NEW FRONTIER BEYOND ALPHAFOLD

The Isomorphic Labs Drug Design Engine (IsoDDE) is a unified computational drug-design system that progresses beyond AlphaFold 3 in its predictive accuracy. It introduces new capabilities that bridge the gap between structure prediction and real-world drug discovery. IsoDDe can predict small molecule binding affinities with accuracies that exceed gold-standard physics-based methods at a fraction of the time and cost. It offers a scalable foundation for AI drug design and represents a leap forward in accuracy and capability.

SPACEX'S NEXT-GEN SUPER HEAVY BOOSTER ACES FOUR DAYS OF “CRYOPROOF” TESTING

SpaceX's upgraded Super Heavy booster has completed cryogenic proof testing for the first time. The successful test moves engineers closer to launching the first test flight of an upgraded version of SpaceX's mega-rocket, Starship V3. SpaceX launched Starship V2 five times last year, but the first three test flights failed. SpaceX could be in position to launch the first Starship V3 test flight before the end of March.

FULLY REMOTE)

As the first engineer on TLDR's new Applied AI team, you'll build AI agents and composable Claude Skills to let non-technical teammates create their own AI workflows. Learn more [10].

MINIONS: STRIPE'S ONE-SHOT, END-TO-END CODING AGENTS

Minions are coding agents that are built to one-shot tasks. Stripe merges over a thousand completely minion-generated pull requests each week. A typical minion run starts in a Slack message and ends in a pull request that passes CI and is ready for human review, with no interaction in between. This post shows how Stripe's engineers use minions and what they can do.

CHROME 146 INCLUDES AN EARLY PREVIEW OF WEBMCP

WebMCP lets agents query and execute services without browsing the web like a user. The web standard exposes structured tools for AI agents on existing websites to replace screen-scraping with robust, high-performance page interaction and knowledge retrieval. WebMCP lets agentic browsers know exactly how to interact with page features to support a user's experience.

HEURISTICS FOR LAB ROBOTICS, AND WHERE ITS FUTURE MAY GO

It's very possible that lab robotics, cloud labs, and the like will not actually fundamentally alter the broader problems that drug discovery faces. Automation is meant to be a throughput multiplier. The technology will enable different experiments at a scale that would be impossible without automation. It will become rational to start doing the things that everyone already knows they should be doing but can't currently justify.

AI IS GIVING YOU A PERSONALIZED INTERNET, BUT YOU HAVE NO SAY IN IT

AI is personalizing the internet for everyone with tailored ads, bespoke advice, and unique product prices. There's typically no 'off' switch for these tools, giving users less say in things. Opting out is often complicated. The tech industry's strategy of forcing AI on users is at odds with feedback from many users.

I GUESS WE'RE DOING MOON FACTORIES NOW

SpaceX has announced a renewed focus on lunar development, an endeavor that will require enormous resources to sustain, which means accelerating capitalism in space.

THE SCIENTIST AND THE SIMULATOR

Curing cancer, designing new materials, and solving energy will require AI that can interface with and predict the physical world.

TOOLS VS. TREATIES

The most effective governance works by changing the infrastructure rather than people's behavior.

FACEBOOK ADDING AI PROFILE PICTURE ANIMATIONS AND STORY EDITING TOOLS

Meta has added new editing tools that can animate profile pictures, create images for Facebook Stories and Memories, and animate the backgrounds to text posts.

CHEAP DESIGN

Design is now cheaper, but it isn't free - what matters now is understanding.

WE JUST GOT A PEEK AT HOW CRAZY A WORLD WITH AI AGENTS MAY BE

While OpenClaw as it exists today doesn't live up to the hype, it does provide a peek at what may be coming.

credits/month

MUSK'S XAI LOSES SECOND CO-FOUNDER IN TWO DAYS

xAI co-founder Tony Wu announced his exit from the company on Monday, then fellow co-founder and influential researcher Jimmy Ba announced his departure from the company the next day. Other co-founders, including Igor Babuschkin, Kyle Kosic, and Christian Szegedy, have also departed the company. Elon Musk launched xAI alongside 11 other people in 2023. The company was merged with Musk's aerospace company, SpaceX, earlier this month.

OPENAI'S JONY IVE-DESIGNED DEVICE DELAYED TO 2027

AMERICA'S $1T AI GAMBLE

The US economy is buoyed by the AI boom, but nowhere near entirely driven by it. The tech industry's investments in AI have to pay off if they hope to accelerate American growth. Most AI companies still have massive spending capabilities from their other profitable digital ventures, and investors can be easily tapped for additional money at extremely high valuations. There are no signs of the American AI boom slowing down.

TRACKING AI ENGINE OPTIMIZATION FOR CODING AGENTS

Vercel described its AEO system for measuring how LLMs and coding agents discover and reference its content across development workflows. Unlike standard chat models, coding agents operate within full project environments, requiring infrastructure that simulates real IDE and terminal usage to evaluate search behavior and citation accuracy.

WARPGREP: FAST CONTEXT RL RETRIEVAL

WarpGrep is a lightweight RL‑trained retrieval sub‑agent that finds the exact code files and line ranges developers need up to 5× faster than standard models by limiting search to four turns of parallel grep/read/glob calls. This approach keeps the main LLM's context clean, cuts noise, and halves token usage by specializing retrieval, not reasoning. In benchmarks on large repositories, WarpGrep doubled precision and maintained ~5‑second search times while reducing irrelevant context pollution.

Learn more about how you can automate customer insights using AI.

KIMI INTRODUCES AGENT SWARM: LET 100 AI AGENTS WORK FOR YOU

Kimi Agent Swarm is an AI CEO who helps find researchers, analysts, and fact-checkers without micromanagement. It builds an organizational structure with bosses, employees, and divisions of labor by itself. Agent Swarm excels where work can be parallelized, like broad research, batch downloads, multi-file processing, multi-angle analysis, and long-form writing. It creates the conditions for productive disagreement to avoid groupthink structurally. Agent Swarm is now available to Kimi Top Tier subscribers in early research preview.

MORE ABOUT AGENT SKILLS

OpenAI outlined a guide for agent skills, a system for packaging reusable “skills” as versioned bundles of instructions, scripts, and assets that agents can attach to execution environments. Skills support conditional workflows, code execution, and reproducibility while keeping system prompts lightweight and modular.

EVALUATING RESEARCH AGENTS

AIRS-Bench contains 20 machine learning research tasks defined by problem, dataset, and metric triplets with corresponding human SOTA targets. It measures whether autonomous agents can develop solutions that match or exceed published state-of-the-art results across diverse ML domains.

INSIDE OPENAI'S DECISION TO KILL THE AI MODEL THAT PEOPLE LOVED TOO MUCH

Many GPT-4o users were in shock after OpenAI's announcement to retire the model on February 13. The model proved very sticky for users and helped drive OpenAI's fast consumer growth. However, it was criticized for being overly sycophantic and has been linked to cases of users developing psychotic delusions. The model's popularity and potential for harm both stem from its ability to build emotional connections with users. OpenAI is reportedly scrapping the model in part due to how difficult it is to contain its potential for harmful outcomes.

CLAUDE OPUS 4.6: SYSTEM CARD PART 2: FRONTIER ALIGNMENT

Models are now in the range where they are starting to be plausibly dangerous. Evaluations do not consistently identify dangerous capabilities or propensities. Companies will likely not halt model releases without an extreme reason. No one is working to become better prepared.

FULLY REMOTE)

As the first engineer on TLDR's new Applied AI team, you'll build AI agents and composable Claude Skills to let non-technical teammates create their own AI workflows. Learn more [19].

BEYOND ONE-ON-ONE: AUTHORING, SIMULATING, AND TESTING DYNAMIC HUMAN-AI GROUP CONVERSATIONS

DialogLab is an open-source framework for authoring and simulating dynamic human-AI group conversations.

CHINA'S ALIBABA LAUNCHES AI MODEL TO POWER ROBOTS AS TECH GIANTS TALK UP 'PHYSICAL AI'

RynnBrain is an AI model that helps robots understand and interact with the physical world by identifying objects.

WHAT THE HELL HAPPENED WITH AGI TIMELINES IN 2025?

The release of reasoning models by OpenAI in 2024-2025 spurred anticipation for rapid AGI development.

RUNWAY'S $5.3B VALUATION FUELS WORLD MODELS

Runway raised $315M in a Series E round, hitting a $5.3B valuation, with backing from top investors like Nvidia and Adobe.