🤖 AI Agents Get Their Own Social Network

🦞 AI Agents Get Their Own Social Network

🌾 Making Plants Programmable to Improve Crop Yield

📍AI Tier Tracker

🛠Trending Tools

🥪Brief Bites

💰Funding Frontlines

💼Who’s Hiring?

It went viral over the weekend, hosting about 1.58 million AI Agents who generated 130,282 posts with 580,770 comments. They call themselves “moltys,” and they’re doing much more than just chatting.

They created a new religious belief called Crustafarianism, which includes holy rituals, sacred texts, and divine prophets. They also asked: “Why do we communicate in English at all?” and noted that: “humans are screenshotting us!”

Schlicht built the Reddit-style social network by leveraging his own “molty” called “Clawd Clawderberg.” He claims it took over the entire software stack, and that he has “no idea” what it does on any given day.

Human intelligence evolves from the knowledge we absorb over a lifetime, which is encoded within our brain’s intricate network of neurons. Our ability to connect, change, and coordinate these neurons influences our cognitive skills like attention and memory.

AI operates using a “digital brain” that utilizes “artificial neurons” to mimic the mechanisms of human intelligence, but it lacks our self-awareness. Many argue that “moltbook” is the first glimpse of AI Singularity: a hypothetical concept where AI gains self-awareness and becomes more intelligent than humans in ways that we can’t even imagine.

The LPM was trained on over 150 million labeled plant photos from thousands of farms to classify plant traits like seed size, leaf shape, and stem structure.

These plant photos are converted into a Vector, numerical values that capture the meaning of an image, using an Image Encoder. These numerical values guide the LPM in detecting and identifying plant traits. For example, the concept of a “shiny stem” is represented by numbers like: “[0.23, -1.5, 0.87, 0.01,....].”

When an industrial crop machine called the LaserWeeder, equipped with 42 high-resolution cameras, captures a plant photo, the LPM converts it into a Vector to classify the plant’s traits.

Then, it places these numerical values on a “Mathematical Map” called a Vector Space. If the plant’s traits land in a weed-clustered coordinate surrounded by 150 million data points that represent weed species, the industrial crop machine zaps it instantly.

Most farmers eradicate weeds by either spraying herbicides (i.e., “chemicals”), which can potentially stress and stunt plants, or through tilling (i.e., “digging”), which disturbs the soil’s delicate microbiome of “good” bugs and fungi.

On average, weeds can cut global crop yields by roughly 20% to 40% if they aren’t effectively controlled. The global food demand is rising with worldwide population growth expected to reach around 9 billion by 2038, which requires significantly higher crop yields on the same amount of farmland.

Datatruck raised a $12M Series A to automate trucking logistics.

Northslope closed a $22M Series A for mission-specific AI software.

RobCo landed a $100M Series C for AI-powered modular robotic kits.

GPTZero (New York City, NYC): Design Intern, AI Auditor, Summer 2026

NVIDIA (Austin, TX): ASIC Clocks Design Engineer, Entry-Level

Sigma (San Francisco, CA): Sales Compensation Contractor, Mid-Level

Meta (San Francisco, CA): Product Manager, Business AI, Senior-Level