💯
Jensen Huang: “AI is not likely to replace you, but someone using AI better than you might.”
That is exactly why AI fluency is becoming a real skill — not just for tech people, but for everyone.
ref:
youtube.com/watch?v=FZh_0uRg…
As a neuroscience researcher building computational tools for multimodal microscopy, Eric Schmidt’s point feels real: coding is shifting from writing every line manually to guiding AI agents that reason, debate, test, and refine code.
youtu.be/I2F2xFvt4mQ?t=671
Tested a microscopy reasoning prompt in napari-chat-assistant (local gemma4:31).
Using structured prompt templates inside the viewer.
Interesting direction for local AI imaging workflows.
github.com/wulinteousa2-hash…
Interesting result 👇
I typed one prompt in napari-chat-assistant and it generated synthetic images, compared histograms, and computed simple SNR live inside napari — not outside the viewer.
A new way to teach image properties interactively.
Try it:
github.com/wulinteousa2-hash…
Quantitation of the physicochemical properties of myelin using Nile Red fluorescence spectroscopy - Teo - Journal of Neurochemistry - Wiley Online Library onlinelibrary.wiley.com/doi/…
Announcing new open-source Python package: aisuite!
This makes it easy for developers to use large language models from multiple providers. When building applications I found it a hassle to integrate with multiple providers. Aisuite lets you pick a "provider:model" just by changing one string, like openai:gpt-4o, anthropic:claude-3-5-sonnet-20241022, ollama:llama3.1:8b, etc.
pip install aisuite
Open-source code with instructions: github.com/andrewyng/aisuite
Thanks to Rohit Prsad, Kevin Solorio, @standsleeping, Jeff Tang and @Johnsanterre for helping build this!
Performing myelin imaging with Nile Red?
Here’s a quick protocol I used:
➡️Prepare 5 mM Nile Red stock solution in DMSO, store at −20°C.
➡️Working concentration: 10-25 μM in 1X PBS.
➡️Stain fixed frozen tissue sections for at least 10 min.
➡️Wash with 1X PBS for 5 min.
1/3
After staining, immerse tissue in PBS on a glass slide. For imaging:
➡️Thin tissue sections: no coverslip needed.
➡️Thick tissues: use a coverslip and secure with a small magnet.
Water-immersion objective lens recommended for optimal imaging.
2/3
Important tips:
➡️Avoid using solvents like xylene, alcohol, or acetone for fixation—they can alter lipid composition.
➡️No detergents or slide mounting media should be used on NR-stained tissue.
#ProtocolSharing#ResearchTips
ref:
pnas.org/doi/abs/10.1073/pna…
Despite non-homogeneous intensity (see red arrow), U-Net achieved 95% accuracy by epoch 10 with augmentation. However, performance dropped by epoch 20 due to overfitting.
2/4
Augmentation, which introduces variation in training data, and early stopping are key to maintaining accuracy, especially when brightness across the tissue varies.
3/4
This study demonstrates that the vibrational spectrum of C-H bonds in lipid molecule's tails, within the myelin sheath encasement, can generate quantum entangled photon pairs through cascade radiation from the second excited state to the ground state. 10/13
The myelin sheath, a lipid membrane encasing the outer side of a neuron's axon, provides energy to the axon, enhances action potential conduction, and acts as an insulator in the nervous system. 8/13
Photons released during tricarboxylic acid cycle in neurons resonantly couple to C-H bond vibrons in lipid molecules, altering possibly the dielectric constant of the membrane to enhance action potential conduction. 7/13
here is GPT-4, our most capable and aligned model yet. it is available today in our API (with a waitlist) and in ChatGPT .
openai.com/research/gpt-4
it is still flawed, still limited, and it still seems more impressive on first use than it does after you spend more time with it.
An AI system defeated elite Chinese doctors in a two-round brain tumor diagnosis competition on both speed and accuracy. This could do incredible good but is another example of areas in which new technology is capable of beating humans. We have to evolve quickly.
The growth model is breaking down in many industries - but we pretend that it is still going. It is corrupting many of us and many of our institutions.