AIb2.io - AI Research Decoded

May 06, 2026

Large Language Models, Jury Duty, and the 900-Paper Pileup

Large Language Models, Jury Duty, and the 900-Paper Pileup

If 12 Angry Men had been set in a systematic review instead of a jury room, you would get something very close to this paper: a stack of 900 studies, several opinionated language models, and a final verdict reached by weighted argument instead of whoever talks loudest.

May 06, 2026

The DNA Potholes Everybody Drives Around

The DNA Potholes Everybody Drives Around

You probably didn't know the same world that gives you phone cameras smart enough to rescue a dim restaurant photo still has a habit of stalling when asked to read a few letters of DNA sitting next to a big genomic fender-bender.

May 06, 2026

The blood was not just sitting there politely

The blood was not just sitting there politely

People headed for rheumatoid arthritis seem to carry a molecular warning signal in their blood years before their joints file the formal complaint.

MAPb2 - 200+ Visual Thinking Tools

Mind maps, flowcharts, whiteboards, and brainstorming canvases - free in your browser.

Explore MAPb2

May 06, 2026

The tattoo that can listen to your muscles and talk back

The tattoo that can listen to your muscles and talk back

What if you could doodle a tiny circuit on your skin, have it eavesdrop on your muscles, and then nudge those same muscles back into action like a coach who lives in your forearm? That sounds like sci-fi with a soldering iron, but it is basically what this paper reports.

May 06, 2026

When the benchmark is a 20-minute sacroiliac MRI, shaving it down to five minutes is not a cute optimization - it is the difference between a clinic running like clockwork and a waiting room slowly turning into a hostage situation.

When the benchmark is a 20-minute sacroiliac MRI, shaving it down to five minutes is not a cute optimization - it is the difference between a clinic running like clockwork and a waiting room slowly turning into a hostage situation.

The study by Deppe and colleagues asks a very practical question: if you suspect axial spondyloarthritis, do you really need the full standard MRI playlist of the sacroiliac joints, or can one high-resolution deep learning-reconstructed DIXON sequence do the job well enough? (Deppe et al.)

May 05, 2026

Boron Learned a New Trick, and the Computer Had Receipts

Boron Learned a New Trick, and the Computer Had Receipts

In the 1950s, Herbert C. Brown gave organic chemistry hydroboration, which was basically boron's breakout role - elegant, useful, and wildly productive. What chemists did not get with that gift basket was a full map of boron's weirder behavior: all the strange reactivity you might unlock if you...

May 05, 2026

Build the interview like it has to survive weather

Build the interview like it has to survive weather

Twenty years ago, researchers tried squeezing future doctors through standard admissions interviews. It didn't work. This paper explains why and fixes it.

May 05, 2026

Combined multi-omics, spectroscopy, and a blood test that might spot glioma without drilling into your skull

Combined multi-omics, spectroscopy, and a blood test that might spot glioma without drilling into your skull

Two types of people - those who already know tiny cellular mail packets can carry cancer clues, and those about to find out that your blood may be gossiping about your brain tumor behind your back.

May 05, 2026

The cell is not one room - it is a whole esports arena

The cell is not one room - it is a whole esports arena

RNA-binding proteins just got caught running a full map rotation across the cell, with 1,768 players tracked by compartment and several reshuffling hard under disease-like stress.

May 05, 2026

The heart's broken playbook

The heart's broken playbook

You can now watch an arrhythmia sweep across an entire mouse heart in 3D and line it up with the tissue that helped cause it, which is a serious upgrade from the old days of trying to understand heart chaos through a few flattened camera angles and sheer optimism. In this paper, Lea Melki and...

May 05, 2026

The microscope slide is a gigapixel gremlin

The microscope slide is a gigapixel gremlin

The design choice that makes this paper click is almost embarrassingly sensible: do not cram a whole pathology slide into one giant model input and pray. Slice the slide into patches, let the model inspect the local evidence, then fuse those clues back into a slide-level verdict. It is the...

May 05, 2026

Too Many AI Papers, Too Few Useful Ones. Then This Lupus Review Shows Up.

Too Many AI Papers, Too Few Useful Ones. Then This Lupus Review Shows Up.

Most AI-in-medicine papers arrive with the same basic promise: give a model a mountain of patient data, shake vigorously, and out pops clarity. Usually what pops out is a PDF and a headache. This one, though, actually does something valuable. Instead of claiming one more shiny prediction tool will...

May 04, 2026

A Moisture Sensor With Better Sight Lines Than Half the Gadgets in Your House

A Moisture Sensor With Better Sight Lines Than Half the Gadgets in Your House

Most touchless interfaces have the architectural grace of a temporary airport kiosk. They work, technically, but they lean on bulky power supplies, short interaction distances, and a general vibe of "please stand exactly here and wave like a confused wizard." This new Science Advances paper by Shen...

May 04, 2026

A Tale of Two Coasts: When AI Maps America's Flood Risk

A Tale of Two Coasts: When AI Maps America's Flood Risk

I'll be honest - when I first saw this paper's title, "A Tale of Two Coasts," I figured it was going to be a straightforward climate doom scroll. Two coastlines, some flood maps, maybe a scary chart. What actually confused me was the machine learning angle. Why would you need three different AI...

May 04, 2026

Robots Are Mixing Chemicals Now, and They're Annoyingly Good at It

Robots Are Mixing Chemicals Now, and They're Annoyingly Good at It

In Isaac Asimov's 1941 short story "Runaround," a robot named Speedy runs circles on Mercury because its programming can't resolve two conflicting directives. Eighty-five years later, researchers have built robots that don't run in circles - they run experiments in circles. Closed loops,...

May 04, 2026

The Case of the Dead Painter's Brushstrokes

The Case of the Dead Painter's Brushstrokes

The clues were hiding in plain sight for four hundred years - microscopic ridges of dried oil paint, each one a fingerprint left at a crime scene nobody knew they were investigating. The suspect: El Greco's own workshop, long accused of finishing The Baptism of Christ after the master died in 1614....

May 04, 2026

The Case of the Missing Follow-Up

The Case of the Missing Follow-Up

Back in 2012, AlexNet made computers weirdly good at recognizing images, but it left one giant hole in the plot: spotting trouble in a picture is not the same as getting an actual patient to walk through the right clinic door afterward. That missing piece is the mystery at the center of this new...

May 04, 2026

The Pocket Is Playing Defense

The Pocket Is Playing Defense

3 reasons this paper matters, starting with the least obvious.

May 04, 2026

The Staging Mystery of Lithium in Graphite

The Staging Mystery of Lithium in Graphite

If your battery were a house, the graphite anode would be the foundation - and for the past thirty years, scientists have been living in it without fully understanding why the basement floods every time they rearrange the furniture. The old approach to modeling how lithium ions nestle between...

May 04, 2026

The buzzer-beater in this paper is pretty wild: a neurology-tuned chatbot came off the bench, took the last shot, and outscored the emergency doctors.

The buzzer-beater in this paper is pretty wild: a neurology-tuned chatbot came off the bench, took the last shot, and outscored the emergency doctors.

That is the basic plot of a new npj Digital Medicine study on Xuanwu-NeuroAid, a domain-specific large language model built for emergency neurological diagnosis. In a prospective shadow evaluation of 433 patients, the model hit 79.4% diagnostic accuracy, while emergency physicians landed at 65.4%....