Bryan Johnson invests $100M in Kernel to unlock the power of the human brain
techcrunch.com
techcrunch.com
Here's someone who's investing his own hard-earned money into a "device aimed at reducing cognitive deficiencies for sufferers of conditions like Alzheimer’s and dementia." And all anyone can do is complain that it's infeasible or that he's unqualified. I don't know about you, but I want more people working on these problems, not less.
Seems very reminiscent of when Musk was starting SpaceX and everyone assumed he didn't know anything about the domain and would automatically fail.
So, lets test your theory by having a look at one of their regularly, successful launches that 14 years investment brought them:
The glory days are over, I think, and I strongly doubt that future generations will keep using the Internet for social interaction.
I have reddit and HN /etc/hosts'd out on my laptops and desktops and mostly just check them occasionally from phone, if you surround yourself with that eventually it starts to soak in.
To be honest with you, if I just look at how those people in my very well developed country (with the Internet being ubiquitous, and being poor generally means receiving minimal benefits as opposed to having no money or place to live) behave who still do partake in the public discourse, I'd say that the problem is more with all of their bitterness than that their bitterness is caused by those who never even had a chance in the first place.
As for me, I'm no saint either because I too can get bitter, annoyed, cynical and depressed from time to time. The difference is just that don't tell people off or post my unhappiness; I have the proclivity to just shut up and leave. "There's no use to arguing". Or I do my absolute best to be constructive, like now.
And to add to that: these days and even years I also barely share my own happiness and/or optimism online, because I don't want my vulnerabilities and/or soft spots to become targets. In fact, I'm not even sharing most of my humanity / personality / identity online with strangers at all anymore, unless it's to reflect or in an attempt to at least offer something to counter all of the now common harshness. Like in comments such as this one. And if I'm really really honest then I also need to admit that this side of me is really closer to a character trait than that it's something implying actual personal involvement.
But I digressed, sorry, so let me back up a bit to the two kinds of groups we've established. Maybe the latter group (the one you pointed out) could indeed bring something nice to the table. Assuming they could, it's not realistic to assume they would, for clearly it's not exactly an hospitable, inviting environment these days, and I'd say this just doesn't make it appealing at all.
Simply put: it is what it is mainly due to those who can join in and they either freeze, fight, or flight. Who in their right mind who's not here already would want to join in on that mess?
The article said that Pfizer annually spends $7-9bn on R&D. Maybe resources are spread too thin; people want moonshots in the biomedical industry as there were in tech, but barriers for entry in this field seem too high.
The benefits are long-term, something I think private investment won't be interested in but perhaps private innovators will be.
The main brain tech guy is Theodore Berger who has tested implants in rats, monkeys and humans with some success http://singularityhub.com/2015/11/15/first-human-tests-of-me...
Berger on video https://www.youtube.com/watch?v=bHubR09oKKE
-VP OF ENGINEERING
-DIRECTOR TO THE CEO
-DATA SYSTEMS ENGINEER / DATA SCIENTIST
-MODELER FOR NEUROPROSTHESIS ALGORITHMS
-MACHINE LEARNING FOR NEUROPROSTHESIS
-SCALABLE NEURAL INTERFACES LEAD (ARCHITECT)
-NEUROSCIENTIST
https://www.scientificamerican.com/article/why-the-human-bra...
...illustrates the problem well. Here's a nice quote:
"Although researchers worldwide were publishing tens of thousands of neuroscience studies every year, neither our understanding of basic brain functions nor our ability to treat brain disorders seemed to be progressing much. "
It's mostly been a failure. The successes have been very, very limited from what I've read. I'm a non-expert that follows generalist scientific publications that report on breakthroughs and interesting products. So, maybe I've missed all the amazing things neuroscience is doing for people's health outside actual surgeries. I just figure spectacular results coming from all the grants and private investment would've made it into things I read. Rarely anything does past "Hmm. That study had some interesting results that might go somewhere some time with a lot more work." Followed by others that seem to contradict it or at least make me wonder.
I remember recently there was an article showing renewed interests in glial cells since they might be more critical to brain function than previously thought. Imagine that. All this research and money thrown at it since the beginning of the field operated on the assumption that they weren't significant. There were others I saw about neural communication in previous years that overturned prior beliefs. One can wonder what new fundamentals we'll learn in this next decade that will challenge our prior beliefs.
Introducing technology that improves how something functions usually needs accurate knowledge about how that something fundamentally works. Investments like Johnson's bet that results will be achieved without that knowledge or even with incorrect knowledge. Kind of like old gold rushes where people bet there'd be gold all over the place in certain areas without even prospecting them first. That's the feeling I get looking at neural ventures.
Note: We had them in the past, too. Implants to control or expand the mind, cause hallucinations, effortlessly give you lucid dreams, and so on. Enough interesting stuff happened that some tactical accomplishments might come from just throwing money at random stuff. Yet, all the interesting stuff of past and present still hasn't given a non-drug guarantee on things as basic as insomnia, depression, or ADD. All failure so far in big picture.
Are you real or a markov chain
Oh no. That's one possible inspiration. Another would be Braintree IPOing or continuing as a company providing an alternative to Paypal's BS I constantly read about while guy continues to make tons of money or get rich. Then he invests in hard problems we're likely to solve where we have the necessary foundation to begin to achieve the stated goals.
Instead, he got rich selling off his good company and now has money to blow on stuff like this. Good for him. Still inspirational with Braintree's early work still something I'm going to cite. Just still worth critiquing for where it went or is going wrong so next bootstrapper considers better alternatives.
Also, realize the irony that you don't like us criticizing what he did with Braintree or where the money is going while Braintree itself owed its success to correct criticism of how successful payment companies were doing things & where their investments were going. He pointed out the problems, showed a better way via his offering, and now we're talking about him instead of them. ;)
That's compatible with my comment. It proposes alternative to the path he took that other bootstrappers with utilitarian bent might find interesting. It doesn't require him personally to do anything. I even said his success was "good for him" in terms of efforts in and result achieved.
"You act like you rose him out of the ashes to save humanity"
You made that up. I explicitly said he didn't. His goal was to get rich. It would take more selfless individuals to do lasting things for humanity. It will take majority of humanity working together to save humanity.
We don't have to guess about his motivations, though. It's in the article:
"For Johnson, the quest is innately personal. His step-father has early signs of Alzheimer’s."
He and his step-father are suffering through the effects of Alzheimer's. He definitely has personal gain from a treatment that would work. There's a clear selfless angle on helping the step-father. He probably, after personally experiencing it, also would like to prevent others from going through the same thing. That's my guess. Respectable motivations for sure.
I guess something like that would allow you to try different computer models to simulate different bit of brain and so decipher its functions bit by bit.
Ummm... that's quite a big leap!
It would be interesting to know how much biology this dude truly explored before plunking down his $100M on some other dude promising him to revolutionize the brain.
>It would be interesting to know how much biology this dude truly explored before plunking down his $100M ...
Not enough.
In practice, collaborations are hard and setting up new ones is even harder, so striking out into the wild and being able to do cool stuff fast generally costs more than 100M.
Estimated timescale: centuries. Not for want of trying, but we have no credible idea what "the" neural code is, even in subcortical areas that are relatively simple to study like the optic and auditory nerves. Once you're anywhere beyond the inferior colliculus, all bets are officially off. (and if you have any good ideas, drop me a line!)
Our best tools fail to explain anything particularly interesting about much, much simpler analogous systems: http://biorxiv.org/content/early/2016/05/26/055624.
https://www.scientificamerican.com/article/why-the-human-bra...
"We built a lunar lander from scratch in seven years."
Hard things get done by people who try to do them. Seriously, it's his own money.
This is a relatively small sum being spent on a project for which very little of the basic science is in place. It may be his own money he's spending, but everyone else is also free to give their honest opinion that he's throwing it down the drain.
No practical comparison between this and Apollo despite me wanting an Apollo-style effort put into certain problems we face. Including understanding the brain although maybe communications infrastructure, cheap computing, and MOOC's first. That will give us more brains for Apollo-style endeavors.
I looked it up - around $20.4B in 1960s dollars, so $160B today. $20B per mooon landing. Not counting Mercury or Gemini.
Nah. Not really Valley Style, lets just fund another AI Moonshot. Because there isn't enough money in AI already.
I mean sure, the ROI investing into the Linux Kernel is 0. But so much money is flowing into AI Moonshots currently that look like their value will be 0 that one gets disheartened quickly when important infrastructure for everyone is needing cash badly, but there is no businessmodel...
I would argue that any GPL-ed software is a hallmark of communism as it is public property and noone is allowed to make any money on it (at least not directly). Therefore, the strongest motivation for hard work (profit) is absent.
Also, it's totally untrue that it's public property (it's private property, just a lot of people's and companies' private property), and that you're not allowed to make money on it "directly". You can make as much money as the market is willing to pay you. If you are selling things that are already publicly available without any added value, that's probably $0, but that's not the result of any permission not being given to you, that's the result of you failing at business. If bottled municipal water is a thriving industry, there's no reason that providing GPL'd Linux patches in a useful form can't be a thriving industry (and in fact it is).
An AI moonshot if it works would change everything. It would change way more than the Linux kernel ever even dreamed of doing.
Ambition isn't the only thing that counts. As much as I'd love to seeing a full cent of every taxdollar going into NASA, handing over the whole dollar would make very people feel very miserable very fast.
I hope this makes the resentment towards the ratio of money flowing into AI Moonshots compared to a little more down to earth things that are needed right now a little more understandable.
I'm a very skeptical and pessimistic man, but can we extend at least some benefit of a doubt here, please.
I have watched Bryan talk before and have been following him. He is determined enough to at least make a significant contribution in this area.
Human intelligence, while fallible and easy to turn off, is arguably our best feature. It's also very versatile.
Bryan Johnson can spend his money any way he likes of course, but one has to wonder why there aren't more initiatives to augment our senses instead of working on intelligence (our own or the intelligence of machines).
I would love to gain a permanent ability to see in the dark, hear ultrasounds, know where North is at all times, etc.
And "all" one has to do with a new sensor is plug it in: the brain is very good at making sense of any new signal, and using it, without prompting or training.
The plugin-in part is non-trivial, of course (although there are shortcuts, such as using the tongue as a crude V0); but one can imagine that once we're able to build a general API to the human brain, many gadgets could be developed for it.
That doesn't sound right. E.g. see stories of blind people who have gotten sight at a late age - I think the tendency is for them not to be able to really utilize sight the way sighted-from-birth people do.
One of the findings after analyzing Einstein's brain was that he had a higher ratio of glias to neurons.
You see, if the brain was simple enough for us to understand its workings, we would not be even capable of understanding the basics of the nervous system functioning (we would have been just too dumb with such simple brains!), and therein lies this irony - We will never be able to fully understand our own brains, because that very impossibility has given us the ability to understand almost everything else in this universe.
So, low complexity results in under-performing intelligent ability. On the other hand, high complexity results in a higher performing brain but more difficult to understand.
I wouldn't necessarily agree with this argument because I suspect intelligent ability is only partially related to neural architecture. Additionally, the ability to perceive things lies in having some kind of framework in which a question can be formed.