Who is Max Hodak?


And why does he say such stupid things?

To answer the first question: he’s a guy with a bachelor’s degree in engineering who developed some software to help colleges find “identify the next generation of high performers” which got bought up by some bigger company, making him relatively rich. Then he hooked up with Elon Musk (Warning! Danger!) to move to Silicon Valley and run Neuralink, despite having no qualifications in biology or neuroscience. He’s a wealthy techbro who has been promoted way above what his competence and experience should allow.

Wait, I think I just answered my second question, too.

New question: why does anyone pay any attention to him? Like me right now for instance?

Because he’s one of the new generation of hucksters whose sole claim to fame is grossly exaggerated promises, and also it helps that he’s associated himself with the crown prince of hucksterdom, Elon Musk. That gets him write-ups in the press, and then we all have to rebut his nonsense, which makes him more of a sensation, which leads to more press, where he gets to spew more nonsense. I don’t know how to get out of this cycle.

I don’t think Twitter helps, either. It enables these bozos to make quick blipverts to promote their idiocy even more. Hodak’s latest was to make this claim:

The co-founder of Elon Musk’s company Neuralink tweeted on Saturday that the startup has the technological advances and savvy to create its own “Jurassic Park.”

“We could probably build Jurassic Park if we wanted to,” Max Hodak tweeted Saturday. “Wouldn’t be genetically authentic dinosaurs but [shrugging emoji]. maybe 15 years of breeding + engineering to get super exotic novel species.”

Pure clickbaity bullshit. No they couldn’t, nor would anyone want them to. Hodak doesn’t even have the expertise to make such a claim, but that’s not going to stop a huckster!

We’re not even close to achieving such a thing, and Musk’s or Hodak’s company doesn’t have the tools to even start. It’s complete, arrogant hype.

Let’s break it down into a simpler problem. Let’s say tigers go extinct (unfortunately, it could happen in our lifetime). Super-rich uber-capitalist gets the fever and decides he’s going to reconstruct the species using “breeding + engineering” to modify house cats, and he gives himself 15 years to do it before his attention span flits off to something equally silly. Can they do it?

No. Maybe someday, but not in 15 years, and that’s a case where we have complete genomic information. Just to mention one obstacle, tigers have a generation time of 8 years. Even assuming the first couple of generations have breeding times of a year, like a housecat, that gives you virtually no time to work out the bugs in your production model. But worse, we don’t have any idea what all the genes that differentiate a housecat from a tiger are! We’re going to need a few decades of work to figure that out, which admittedly, would be an interesting research program, but doing it with the goal of making a tiger would be unproductive, especially given that we don’t seem to be able to keep the existing tiger population alive.

And that’s the easy problem, compared to resurrecting dinosaurs. The only templates we have for the dinosaur genome have been extensively modified by over 70 million years of drift and selection, and we don’t know what genes were lost or gained, or what their role in the complex outcome of “dinosaur” might be. It would be lovely to find out, but it’s not the accomplished fact Hodak thinks it is.

Also, “Jurassic Park” is fiction, based on a bad novel written by a hack writer of thrillers. I read it when it first came out, as an undergrad who was waffling between an oceanography and biology major, and it’s one of the first novels I recall ripping up and throwing in the trash because the science was so bad. It’s kind of a shame that it got rescued by CGI and movies.

As for the Neuralink connection, which I’ve written about a couple of times, that’s also bad science. The plan is to build a brain-machine interface, so you can just think at a computer and have your brilliance manifest in code, or control a fighter plane even faster. Here, though, is their great accomplishment:

Launched in 2017, Neuralink works on creating brain-computer interfaces with the hopes to one day help those afflicted with Alzheimer’s disease, dementia, paralysis and spinal cord injuries, among others.

In August 2020, Musk debuted Gertrude, a pig that Neuralink had implanted a small computer chip in its brain. The chip was planted near the part of the brain that controls its snout, so as Gertrude ate, a computer showed waves and spikes being emitted from the chip, monitoring Gertrude’s neural response.

Uh, right. My first year in grad school we repeated a classic experiment, placing hook electrodes on a cricket’s cercal nerve, and recording the pattern of “waves and spikes” as the insect processed sensory information. It was cool — you could see differences when you touched or moved the cerci, or with blowing on them from different directions, and this is just more of the same, only they’ve got it on a chip rather than the boxy little pre-amps and clumsy oscilloscopes we used. Congratulations. They’re catching up with JZ Young, who was doing recordings of neural activity 70 years ago.

One other difference: Young and that generation of neurophysiologists were working on organisms acutely — the animals were dead after the experiment was over. Are you willing to have a chip inserted in your motor cortex so you can play video games better?

The most biting sentence in the article is this one.

Hodak didn’t further explain what technology Neuralink could use to engineer the long-extinct dinosaurs.

Exactly. Hodak is talking out of ignorance, nothing more. Don’t listen to him.

Comments

  1. specialffrog says

    If the Boring Company is any indication, they will declare victory after breeding slightly large orange cats.

  2. hemidactylus says

    We could atavize a crow to become a really smart dinosaur (it already is). Give me billions.

    Though a little suspect, the notion of neurofeedback has been using “brain waves” to manipulate cursor or arrow position on a computer game output for decades. Low tech compared to chip interfaces. “They” were able to get a low quality rendition of what someone was watching via brainscan transformation magic already without the invasiveness.

    https://abcnews.go.com/Health/MindMoodNews/scientists-youtube-videos-mind/story?id=14573442

    Ok an fMRI is expensive and unwieldy. Still just somehow reverse the process.

    I don’t discount a potential future in brain-machine interfacing, but just expand on the biofeedback model a bit. Maybe some version of the God-awful helmet reversed. Voice recognition commands seem more realistic.

    Invasive brain prosthetics could be something for people needing to overcome some physical issue, but not for Minecraft or whatever the kids are playing. Just keep it out of Musk’s hands.

  3. raven says

    And that’s the easy problem, compared to resurrecting dinosaurs. The only templates we have for the dinosaur genome have been extensively modified by over 70 million years of drift and selection, and we don’t know what genes were lost or gained,…

    The dinosaur templates we do have are obvious, the flying dinosaur template, birds.
    And yeah, we are a long way from being able to modify a chicken into a dinosaur again.

    Science Mutant Chickens Grow Teeth
    Feb. 21, 2006 , 12:00 AM

    Warning: Mutant chickens may bite. Researchers have identified a genetic mutation that creates incipient teeth in bird embryos. The discovery provides a modern day glimpse of a feature that hasn’t been seen in avians for millions of years.

    Birds lost their choppers 70 million to 80 million years ago. That’s what made an experiment in 1980 so surprising: After scientists grafted oral tissue from mice onto a chicken’s gums, the birds grew round, mouselike teeth.

    Chickens still seem to have most of the genes to make teeth, even though they have beaks instead of teeth.
    This mutation is ultimately lethal though. “…talpid2–because embryos with the mutation rarely survive past 12 days.”

    To make an imitation dinosaur out of a chicken would take a deep understanding of the genetics of developmental biology and a lot of empirical research. Any time I would pick would just be a wild guess, but many decades at least.
    And then the question would be why? It would cost a lot of money and we have more important things to deal with. Right now we are fighting our way out of a newly evolved virus pandemic.

  4. PaulBC says

    “We could probably build Jurassic Park if we wanted to,” Max Hodak tweeted Saturday. “Wouldn’t be genetically authentic dinosaurs but [shrugging emoji]. maybe 15 years of breeding + engineering to get super exotic novel species.”

    This sounds like the beginning of the old Maxwell Smart gag “Would you believe?”

    Would you believe? Real live dinosaurs!
    No.
    Would you believe ‘super exotic novel species.’?
    No.
    How about hamsters in little tutus?

  5. raven says

    Published online 31 October 2006 | Nature | doi:10.1038/news061030-4 Edited for Length
    Ancient human virus resurrected
    Virus from distant past may throw light on role of retroviruses in cancer. Helen Pearson

    Researchers in France have recreated a 5-million-year-old virus whose remains are now found littered across the human genome.

    The virus is of a type called a retrovirus, which can insert copies of its genetic material into our own DNA. The relics of these copies in human DNA are called human endogenous retroviruses, or HERVs.

    Today, copies of HERVs make up some 8% of our genetic code. But these copies have racked up mutations and are largely obsolete: scientists have never found one that can still convert itself into new, infectious virus particles.

    This is the state of the art for resurrecting extinct organisms.
    We’ve done it twice with two different viruses.
    The 1918 Influenza virus has been resurrected.
    We have also resurrected a human retrovirus that has been dead for 5 million years.

    Needless to say, it is a long way in complexity between a few viruses and a dinosaur.

  6. hemidactylus says

    Is it possible Neuralink could use neural prosthetics and genetic engineering to turn Elon Musk into a person? Or at least less of a prick?

  7. PaulBC says

    15 years of gazillionaire funding probably could accelerate the advance of medical technology. A team at UCSF has been working on an artificial implantable kidney for a long time. Nothing about it looks impossible, but what strikes me is what a tiny, shoestring budget it has compared to the things that do get funded by venture capital around here, that sometimes succeed financially, fail more often, and almost never improve human life in a meaningful way.

    I can’t even imagine why “Jurassic Park” (or landing on Mars) would pop into anyone’s head first when trying to figure out what to do with unlimited budget and expertise.

  8. PaulBC says

    I read it when it first came out, as an undergrad who was waffling between an oceanography and biology major, and it’s one of the first novels I recall ripping up and throwing in the trash because the science was so bad. It’s kind of a shame that it got rescued by CGI and movies.

    Jurassic Park came out as a novel in 1990, so I think you have the chronology wrong. I remember reading some discussion about it on Usenet before the movie was even on anyone’s radar. I was in grad school and I am younger.

  9. says

    What would resurrected dinos eat? I assume a large percentage of what plant eaters ate is probably extinct. Would meat eaters be able to digest the meat of modern creatures?

  10. says

    Oh, yeah, now that I think of it — I think I picked up Jurassic Park as airplane reading while I was doing postdoc interviews. So I had a bit more scientific background to see how bad it was.

  11. hemidactylus says

    Ok blame pessimistic Frankfurter critique for warping my brain, but we already instrumentalize and subjugate people and “people” as objects are already made mere adjuncts to machines if not superseded altogether. What happens when you integrate the machines physically into the humans? Cui bono?

    And alas sex it up with the techbro angle. Lure them in with the trappings of a better future (for some). And resurrected dinosaurs. What could go wrong?

    But thankfully our hubris will result in Transformer dinobots that when unleashed will in a completely unforeseen fashion end our dystopian misery forever. Dinobots don’t care to exploit. They just chomp everything in sight to bits because they can.

  12. consciousness razor says

    But why would we want to put all of our high-performing dinodroid college graduates on Mars? That’s the real question. If we want America® to remain competitive in the 22nd century and beyond, we should be keeping them here to enhance* our workforce and embiggen its productivity.**

    *replace
    **corporate profits

  13. PaulBC says

    “Max Hordak” would be a great name for another defective clone of Horde Prime if they do a new season of She-Ra. (Actually, what if it was a glitchy computer simulation of Hordak like Max Headroom. The possibilities abound.) (And yes, I am a nerd, why do you ask?)

  14. microraptor says

    I’m reminded of Heck cattle- the 100 year long attempt to breed modern cattle into aurochs that has resulted in an animal that looks and acts nothing like a real aurochs.

  15. KG says

    the animals were dead after the experiment was over. Are you willing to have a chip inserted in your motor cortex so you can play video games better?

    I’m not disagreeing with your opinion of Hodak, or Musk, but as I understand it, Gertrude the pig wasn’t dead. And no, I wouldn’t agree to chip insertion so I can play video games better (aside from anything else, I don’t play video games), but OTOH, if a stroke had left me largely paralysed…

  16. stroppy says

    I’d be less bothered by some frivolous fiction, but Michael Crichton was also a pernicious AGW denier. What a milieu these bobble-headed influencers have constructed for themselves.

  17. davidc1 says

    As Homer J Simpson would say “Jurassic Park ,They tried that in the movies ,and it didn’t work.”

  18. consciousness razor says

    I guess I sometimes forget how ancient I am.

    For reference, the Jurassic Park book came out during season 4 of Star Trek: The Next Generation (more precisely, the day after episode 9, Final Mission, referring to Wesley Crusher who “will be missed”).

    Also, the Game Boy had been released only a few bitterly short months earlier (for Japan and North America, but not yet Europe or South Korea). Tetris was pretty fun.

    You are indeed old.

  19. Jado says

    When these sorts of hucksters speak gibberish in public, what do you think the reaction is of actual serious academic researchers in their companies (assuming there ARE people like this in the companies – could be just a circle of yes-men)? Do you think the people with expertise are tearing their hair out at what their idiot boss just uttered, or do you think they are cynical enough to await the inevitable influx of enthusiastic moron venture capital dollars to then talk about budgets and bonuses, like common Star Wars defense contractors in the late eighties negotiating with a dementia-addled Reagan? I gotta think anyone still employed by such obvious grifters falls into the later category.

  20. consciousness razor says

    Sorry, I was looking at the wrong dates…. The Game Boy was actually released in Europe about one month before Jurassic Park, and in the same month in South Korea (Nov. 1990). By then, it had been selling for a little over a year in Japan and North America.

    I didn’t have one. I was always more into computer games. But Tetris was pretty fun.

  21. PaulBC says

    Jado@24

    When these sorts of hucksters speak gibberish in public, what do you think the reaction is of actual serious academic researchers in their companies (assuming there ARE people like this in the companies – could be just a circle of yes-men)?

    It depends on the company. Theranos was able to attract “big names” to its board such as former Secretary of State George Shultz, but these people had no ability to evaluate Elizabeth Holmes’s claims. Her idea had been flagged as unrealistic by a Stanford professor, and some investors (Google ventures) steered clear. (That’s what I recall and I did not verify just now.) So what was the “adult supervision” scientifically speaking? By the time they were sending samples to outside labs, they were running a scam, but maybe they had some dreamers who thought they just needed a little more time to get it to work. I am not sure. If you have the investment capital, you can always put together a company that will claim to be doing something or other. How much is pure deception and how much is self-deception will vary from case to case.

  22. unclefrogy says

    @7
    that would be my thoughts as well.
    though the idea of controlling prosthetics limbs with some kind of brain interlink would be a useful accomplishment

    as far as chicken teeth go teeth are well and good but making or resurrecting a jaw for them might be a little harder
    uncle frogy

  23. PaulBC says

    @7 @28 Maybe Musk could be tricked into getting into some kind of near-fatal accident and then reconstructed as a cyborg like General Grievous. It would be win-win since Elon Musk would probably think it was cool. (OK, really just “win” because I am not sure what the point of it would be besides that.) Done babbling. Carry on.

  24. consciousness razor says

    Maybe Musk could be tricked into getting into some kind of near-fatal accident and then reconstructed as a cyborg like General Grievous.

    But rather than commanding the Confederacy’s army, he could play Pong without relying on hand-eye coordination.

    Strangely, it still sounds like something that asshole might want to spend a few billion dollars on.

  25. birgerjohansson says

    Heck cattle…The wild aurochs disappeared in historical times, so I think Svante Pääbo and his team could provide an authentic auroch reference genome, if someone financed it.
    Mammoths, not so much (unless quantum computers might re-assemble those highly fragmented genes).
    Also, smilodon. If we are doing mammoths and wooly rhinos, I demand we do the sabre-toothed kitties as well.
    .
    OT For the first time, a high-fidelity map of human chromisone 8 has been assembled, including things like highly repetitive DNA that older tech could not assemble.

  26. hackerguitar says

    The promise of machine learning is real, but overconfident knuckleheads like this gent make it – a serious tool for training behaviors in very specific domains – look ridiculous.

  27. lucifersbike says

    hemidactylus@2; “We could atavize a crow to become a really smart dinosaur (it already is). ” There are large numbers of four different corvid species living in my street. I think they’re just fine as modern representatives of the archosaurs; especially the magpies that enjoy tormenting the local cats by keeping just out of reach. It might be more worthwhile to edit the human genome so that our descendants think before they choose to follow arseholes.

  28. stroppy says

    So I’m out of my depth here, but for purposes of rewilding, heck cattle (or whatever) wouldn’t need to be exact replicas of aurochs, just sufficiently able to fill the role of natural grazers that are now missing from the landscape–not that I know what ‘sufficient’ in this context is. In the US, one of the environmental arguments for maintaining some form of cattle ranching is that it can fill the niche of extirpated grazers.

    Auroch genome:
    https://genomebiology.biomedcentral.com/articles/10.1186/s13059-015-0790-2

  29. PaulBC says

    stroppy@35 According to Wikipedia:

    In the first years of the Weimar Republic, the brothers Heinz and Lutz Heck independently started their extensive breeding-back programmes. Their motivation behind that was to rescue the aurochs from oblivion because it was constantly confused with the wisent, the other large bovine of Holocene Europe. The Heck brothers believed that creating a look-alike and showing both species next to each other would help to show the difference between the two species to a broader public.

    …which is the most weirdly specific reason I have ever heard for doing anything. I guess animatronics weren’t available back then. (How about scale models in clay?) So anyway, an animal that just fit the same niche would not have suited the above purpose.

    Considering their only tool was breeding back, I suspect there are many more possibilities now if you really want to bring back aurochs. I would not pretend to set a timeline for it though.

  30. stroppy says

    Indeed. The history of genetics seems to be peppered with weirdness.

    These days much of the interest in aurochs is part of the move to rewild Europe and restore ecosystems to the extent possible. It seems to have become one of the charismatic mascots for that project, which may be a double edged sword in terms of how people perceive it and the science behind it.

    PBS series with the larger context
    Europe’s New Wild
    https://www.pbs.org/video/preview-04pncv/

    Rewilding Europe
    Auroch
    https://rewildingeurope.com/search/auroch/

    Wild aurochs-like cattle reintroduced in Czech Republic (Update)
    https://phys.org/news/2015-10-wild-aurochs-like-cattle-reintroduced-czech.html

    The wild, extinct supercow returning to Europe
    https://www.cnn.com/2017/01/09/world/auroch-rewilding/index.html

    etc.

  31. says

    It’s refreshing to see skepticism from knowledgeable experts of half k ideas promoted by cult like popular figures like Elon Musk. Companies like Neuralink are pretty much a novelty and even with some success aren’t going to have much effect on most peoples lives. Musk is also a big promoter of solar and batteries for the electrical grid and this is where he should be getting more scrutiny from knowledgeable experts like physicist Mark Mills:

    https://www.manhattan-institute.org/green-energy-revolution-near-impossible

    I think this is diverting resources from real proven grid solutions like nuclear power.

Leave a Reply