After each natural action of going to the toilet, flushing the toilet seems like an insignificant act, but for researchers at the Massachusetts Institute of Technology (MIT) in the United States, it is equivalent to flushing away a treasure trove of data.
Recently, researchers at MIT have partnered with Seed Microbiome, a microbiome company, to use the newly developed Auggi artificial intelligence system to analyze photos of people's feces. The researchers hope to collect fecal photos from 100,000 volunteers and establish a human fecal database to analyze the participants' health conditions based on the fecal samples.
In the United States, approximately one-fifth of the population suffers from chronic gastrointestinal diseases such as Crohn's disease. Gastrointestinal diseases can change the color and structure of human feces, so analyzing the color of feces can help patients control their condition.
To achieve this goal, Seed Microbiome needs a large number of fecal samples. They are hoping to collect 100,000 fecal images from around the world and use AI technology to analyze these images to identify which ones are normal and which ones have abnormalities.
Seed Microbiome states that people often overlook the important indicators of their overall health, such as the size, shape, color, composition, concentration, and frequency of bowel movements.
The condition of the human digestive system is often overlooked, and the research team hopes to gain a deeper understanding of the human digestive system through these studies. They estimate that three-quarters of the global population have some form of digestive problem, such as constipation, diarrhea, bloating, or irregular bowel movements.
Professional doctors will carefully examine each received image. Stool can be classified into seven categories according to the Bristol Stool Scale, which can indicate whether a person has constipation or lacks fiber, among other conditions. The insights of doctors regarding stool will help train artificial intelligence models to understand the same content seen by doctors in the images.
According to David Hachuel, co-founder of Auggi, the start-up company building the platform, similar training systems are used to teach self-driving cars how to recognize trees or cats on the road.
Hachuel stated that the goal of this movement is to eventually make the database an open-source tool for academic researchers. At the same time, they hope to empower patients to regain control of their intestinal health. He stated that they are developing technology to help people with chronic gastrointestinal diseases better understand the relationship between their lifestyle and symptoms.
If we are unwilling to provide fecal images for their research, how can we understand our own intestinal condition through feces? The correct answer is: go to Japan and buy a toilet!
LIXIL, the largest building materials and equipment group in Japan, can sell toilets worldwide. In October of this year, LIXIL made a big move and released an AI smart toilet equipped with a visual detection system.
For example, one of LIXIL's toilet series, SATIS, has a built-in exclusive application that can record the time and status of each toilet visit and view the water and electricity usage for each toilet visit.
According to the introduction, this AI toilet can automatically classify the shape and size of feces according to international standards. If nursing institutions use this toilet, staff can understand the residents' physical condition from the analysis of their feces, thereby helping them manage their health.
The AI toilet is equipped with a sensor camera inside. When the user defecates, the toilet immediately performs image recognition on the feces, dividing the shape into seven stages and the size into three stages.
The data obtained from AI recognition is uploaded to the administrator's personal computer through the network.
The judgment criteria for feces refer to international standards and are being improved in collaboration with relevant doctors.
From an experiential perspective, the smaller and thinner the feces, the higher the possibility of constipation and diarrhea. Therefore, staff can determine whether users are troubled by constipation or diarrhea just by looking at the results of fecal analysis.
According to official statements, during the development process, approximately 3,000 images were collected in collaboration with employees, and the accuracy of the toilet's recognition may be as high as 80%.
Of course, this special toilet is naturally targeted at special users, mainly for the elderly who require care or patients with dementia, helping medical staff better judge the physical condition of these users.
In the past, medical staff could only visually record and judge the bowel movements of these elderly patients. The AI toilet simplifies this process further.
It is both reassuring and disappointing that it has not yet been determined whether this toilet will be commercially produced. The development of this toilet took four years, and a demonstration experiment will be conducted in 2020 to determine its practicality.
Finally, if you want to help establish the world's first fecal database, please visit seed.com/poop on your mobile phone (because it is strange to bring a laptop to the toilet, so this page does not allow you to submit photos unless you use a mobile phone).
Then, click on the purple button that says "#GIVEaSHIT". The system will prompt you to enter your email address and the schedule of your bowel movements in the morning, afternoon, or evening. After you submit your fecal photo, the image will be separated from the metadata (such as your email address and other potential identifying information) to ensure anonymous donation.
Thank you for your generous contribution.
Alphabet Q1 paid subscribers reached 350M, fueled by YouTube & Google One. YouTube ad revenue slightly missed at $9.88B amid Premium growth. Total Q1 revenue hit $109.9B, beating forecasts.
Anthropic releases Claude connectors for Adobe, Blender, and Ableton, enabling AI workflow automation and app control. It also sponsors the Blender Development Foundation.
Apple introduces an App Store model splitting 12-month plans into monthly installments. On iOS 26.4+, it lowers upfront costs without reducing total price, adding billing tracking.
Google launches Aadhaar digital IDs in Wallet for India, enabling secure QR/NFC verification. A phased partner rollout targets key sectors, while passport-based credentials expand globally.
YouTube tests "Ask YouTube," an AI tool blending text and video for complex queries. Initially for US Premium users, it supports interactive follow-ups, expanding Google's AI search.
Google grants US DoD classified AI access, joining OpenAI and xAI in defense deals amid Anthropic's dispute. Contractual safeguards against surveillance and lethal weapons remain unenforceable.
Mistral AI launches Workflows, a Temporal-powered orchestration platform for secure, Python-driven enterprise AI automation featuring human-in-the-loop approvals.