DSC-RX100M3
ƒ/1.8
8.8 mm
1/30
160

At the start of the live update event today: full video and cyber-pig ping test.

It felt like a scene from a sci-fi movie when we entered.. surrounded by brain robots, and the sounds of neural-jacked pigs rustling behind the curtains…

C|Net: “Elon Musk shows Neuralink brain implant working in a pig. A wireless link from the Neuralink computing device showed the pig’s brain activity activity as it snuffled around a pen on stage Friday night. The demonstration shows the technology to be significantly closer to delivering on Musk’s radical ambitions than during a 2019 product debut, when Neuralink only showed photos of a rat with a Neuralink connected via a USB-C port. The FDA granted approval for “breakthrough device” testing.

“It’s like a Fitbit in your skull with tiny wires,” Musk said of the device.

It communicates with brain cells with 1,024 thin electrodes that penetrate the outer layer of the brain. Then there’s a Bluetooth link to an outside computing device, though the company is looking at other radio technology it can use to dramatically increase the number of data links.

Neuralink has a medical focus to start, like helping people deal with brain and spinal cord injuries or congenital defects. The technology could, for example, help paraplegics who’ve lost the ability to move or sense because of spinal cord injury, and the first human uses will aim to improve conditions like paraplegia or tetraplegia.

“If you can sense what people want to do with their limbs, you can do a second implant where the spinal injury occurred and create a neural shunt,” Musk said. “I’m confident in the long term it’ll be possible to restore somebody’s full body motion.”

“The future is going to be weird,” Musk said, discussing sci-fi uses of Neuralink. “In the future you will be able to save and replay memories,” he said. “You could basically store your memories as a backup and restore the memories. You could potentially download them into a new body or into a robot body.”

Musk also discussed seeing in infrared, ultraviolet or X-ray using digital camera data. “Over time we could give somebody super vision,” Musk said.”

Full disclosure: Future Ventures is an investor

7 responses to “Elon Musk and the Neuralink Future”

  1. Looks like the robot is sewing my brain heading inOne benefit of using pigs as a test animal; they can be trained to walk on a treadmill. And with the wireless Neuralink puck implanted, you can see how they can read out the pig’s proprioception of its front leg position (faint lines) compared to the actual position (dark lines recorded by motion capture of the BGRP color spots indicated on the far right). The brain compensates for signaling lag effects by thinking the leg is in a certain position slightly before it is. Jump to this segment of the video

  2. Neuralink President Max Hodak wrote a recent blog post on AGI futures: "I came to view consciousness and intelligence as distinct phenomena. Both are rare, but while one is precious and fragile, the other is one of the universe’s great hazards. If we know that information flowing through sequences of configurable nonlinearities is expressive enough to produce general intelligence, we also have a well-known algorithm for designing those networks: evolution. If you zoom out, fundamentally we know that with enough compute power, genetic search is capable of designing neural networks that are at least as smart as humans. Where I begin to worry about AGI with independent agency is when we start to talk about optimizing them under natural selection. My hypothesis is that when you select for survival rather than artificially selecting for some other property you enter different territory, and particularly I expect this is where you begin to see violence. There is nothing special about doing this; it is just a different software environment, and banning it effectively is impossible. With advances in compute power, the ability to fit such networks will be widespread well within a decade. Faced with autonomous artificial intelligence, developed under a selection for its survival, humans will immediately be in a precarious situation."

  3. I shared some remarkably similar concerns in 2004, my first year of blogging:

    is the desire for self-preservation coupled to intelligence or to evolutionary dynamics?… or to biological evolution per se? Self-preservation may be some low-level reflex that emerges in the evolutionary environment of biological reproduction. It may be uncoupled from intelligence. But, will it emerge in any intelligence that we grow through evolutionary algorithms?

    Given the iterated selection tests of any evolutionary process, is it possible to evolve an intelligence without an embedded survival instinct?

    And in the comments:
    Will the equivalent of the “reptilian brain” arise at the deepest level in any design accumulation over billions of competitive survival tests?

    A related question: can the frontier of complexity be pushed by any static selection criteria, or will it require a co-evolutionary development process?

    Yudkowsky replied:

    “This problem is intrinsic to any optimization process that makes probabilistic optimizations, or optimizes on the basis of correlation with the optimization criterion. The original criterion will not be faithfully preserved. This problem is intrinsic to natural selection and directed evolution.

    Building Friendly AI requires optimization processes that self-modify using deductive abstract reasoning (P ~ 100%) to write new code that preserves their current optimization target."

    I don’t think self-modifying code is possible in neural networks. And thus, there is no hard take off. We will need to train a new network to make it smarter, not have it self-modify its hyperparameters. We will build an intelligence that exceeds human intelligence before we reverse engineer our own brain, or any AI brain of comparable complexity.

    The debate then ropes in Rosedale and Drexler. Wild to revisit this.

  4. Max also observes "Simply detecting that an unsupervised AGI is on the internet is non-trivial." For sure. Reminds me of my most popular tweet to date, dashed off during a dinner with Richard Garriott:

    A sufficiently advanced AI would chose to fail the Turing Test.

  5. Saw the live show. Super dope.
    Question: why do you continue to use flickr?

  6. 16 years of tagged archives… and easy discovery and re-use by Google, Wikimedia, Books under creative commons license. Some examples

  7. and now, for example, this is the wikipedia photo: en.wikipedia.org/wiki/Elon_Musk#Neuralink

Leave a Reply

Your email address will not be published. Required fields are marked *