The Neuralink part is so cutting edge. I be it’s just a matter of time and red tape until they use the chip to talk to the exoskeleton units that exist to allow for people to use them using the chip.
I similarly wonder if there is an opportunity to bypass the spinal cord injury and communicate to the neurons directly below. Going full Gundam is a fun alternative though!
Not trying to start a debate but it’s really sad that people (including the media) were really trying to say the whole test failed because of the hiccups. Why wouldn’t they be for private healthcare essentially providing amazing technology to increase the functionality within their lives. This technology is breathtaking.
It's a Phase I trial, so there were bound to be some challenges translating this to humans. However, the cause of the threads retracting is believed to involve air left in the skull after the surgery. This isn't dangerous and is normally reabsorbed by the body. Fortunately, the Neuralink engineers developed a workaround to maintain high performance.
As for your question about the negative press - Musk is perceived politically. The journalists writing for these outlets can't contain their bias.
To take one of the many, many technological advances that are way over my head, I don't understand AlphaFold 3, or any AI modeling applied that way. If enough is known about protein structures and behaviors to program the AI to model and predict outcomes, then why not just save a step and skip the AI? Isn't AI still a case of GIGO, too? How are its predictions verified in real life? If they are verified in real life, then why insert software into the equation at all? I'm not saying that the predictions aren't an awesome thing, but can't the humans who wrote the software and input all its "knowledge" do the same thing the software's doing, since they had the information in the first place? Isn't that what chemists already do: predict molecular interactions based on the known properties of the molecules? Does it risk degrading chemists' skills the way automation has degraded the skills of pilots who didn't come along in the stick-and-rudder era?
On the other hand, I can't remember the last time I learned about something as cool as Regeneron. Has there been resistance to that the way there was to cochlear implants, as damaging to "deaf culture"?
Great questions! Predicting 3D structures from amino acid sequences is incredibly complex—we don’t know how to program it directly. So how do we figure out how a string of amino acids will fold into a 3D shape? AlphaFold 3 is trained on enormous amounts of data to make these predictions. By iteratively performing this on structures with validated experimental data, such as those obtained through techniques like X-ray crystallography, it learns patterns we would never see as humans.
These tools are extremely valuable because we can then build novel proteins designed to bind to target proteins in the body. It’s essentially an emerging field of in silico medicine. While it still requires validation in a laboratory, it has the potential to significantly expedite the rate of development.
Regeneron is doing some interesting pioneering work. I lived in upstate NY for a while and had several friends working at their facilities. These products are still in their infancy for restoring hearing, and I’m sure when they become more viable, they will spark controversy. I also wonder how long it takes for the brain to rewire to understand and interpret sound.
The Neuralink part is so cutting edge. I be it’s just a matter of time and red tape until they use the chip to talk to the exoskeleton units that exist to allow for people to use them using the chip.
I similarly wonder if there is an opportunity to bypass the spinal cord injury and communicate to the neurons directly below. Going full Gundam is a fun alternative though!
Not trying to start a debate but it’s really sad that people (including the media) were really trying to say the whole test failed because of the hiccups. Why wouldn’t they be for private healthcare essentially providing amazing technology to increase the functionality within their lives. This technology is breathtaking.
It's a Phase I trial, so there were bound to be some challenges translating this to humans. However, the cause of the threads retracting is believed to involve air left in the skull after the surgery. This isn't dangerous and is normally reabsorbed by the body. Fortunately, the Neuralink engineers developed a workaround to maintain high performance.
As for your question about the negative press - Musk is perceived politically. The journalists writing for these outlets can't contain their bias.
Sorry I meant functionality in people who are critically disabled’s lives.
Super impressive stuff. The Neuralink is amazing progress, as is AlphaFold's update. We are living in the future!
It's a fun time to be alive! So much opportunity - I think there is a lot of opportunity to work on translational products with alphafold.
To take one of the many, many technological advances that are way over my head, I don't understand AlphaFold 3, or any AI modeling applied that way. If enough is known about protein structures and behaviors to program the AI to model and predict outcomes, then why not just save a step and skip the AI? Isn't AI still a case of GIGO, too? How are its predictions verified in real life? If they are verified in real life, then why insert software into the equation at all? I'm not saying that the predictions aren't an awesome thing, but can't the humans who wrote the software and input all its "knowledge" do the same thing the software's doing, since they had the information in the first place? Isn't that what chemists already do: predict molecular interactions based on the known properties of the molecules? Does it risk degrading chemists' skills the way automation has degraded the skills of pilots who didn't come along in the stick-and-rudder era?
On the other hand, I can't remember the last time I learned about something as cool as Regeneron. Has there been resistance to that the way there was to cochlear implants, as damaging to "deaf culture"?
Great questions! Predicting 3D structures from amino acid sequences is incredibly complex—we don’t know how to program it directly. So how do we figure out how a string of amino acids will fold into a 3D shape? AlphaFold 3 is trained on enormous amounts of data to make these predictions. By iteratively performing this on structures with validated experimental data, such as those obtained through techniques like X-ray crystallography, it learns patterns we would never see as humans.
These tools are extremely valuable because we can then build novel proteins designed to bind to target proteins in the body. It’s essentially an emerging field of in silico medicine. While it still requires validation in a laboratory, it has the potential to significantly expedite the rate of development.
Regeneron is doing some interesting pioneering work. I lived in upstate NY for a while and had several friends working at their facilities. These products are still in their infancy for restoring hearing, and I’m sure when they become more viable, they will spark controversy. I also wonder how long it takes for the brain to rewire to understand and interpret sound.