We Can Stuff Zetabytes of Data into DNA (Someday)

Links:
The Asianometry Newsletter: www.asianometr...
Patreon: / asianometry
Threads: www.threads.ne...
Twitter: / asianometry

[@hupekyser]

All your base pairs are belong to us

[@johnsmith1926]

Lab jargon is actually just bases... 🙂

[@schilling3003]

I love how all of his videos include jokes like this. The delivery is so dry, I'm sure most people completely miss most of them. I'm sure I have missed many of them.

[@dlmhdlmh]

Love it

[@comosaycomosah]

Based pair

[@BitchinSpectre]

I get the reference

[@ChuckSwiger]

Was always surprised the human genome project never ran across a sequence that could be interpreted as something like, "Copyright Andromeda Planet Seeding Corp."

[@obsidianjane4413]

Would we be able to recognize it even if there were?

[@privacyvalued4134]

There was an episode of Star Trek TNG on this: Season 6, Episode 20 "The Chase." We just need to get to Indri VIII before the Klingons do.

[@JO_______]

Would essentially be impossible for a number of reasons. A) If the planet was 'seeded' for life, the dna would have mutated far past the point of recognition. We're talking a few billion years of evolution at this point. That alone should be a problem even attempting to verify the hypothesis, but here's a couple more reasons: B) We have no way of telling random noise from a nonhuman civilization in dna. In order to 'store' data, we have to establish what each 'bit' means, and that's an interpretive thing. C) DNA decays pretty easily outside a body in uncontrolled environments. It's why the 'mosquitos in amber with dino dna' aspect of jurassic park is fictional. It doesn't actually work in reality past a few thousand years unless the dna is stored in a specific environment that can preserve it. (amber ain't that environment). Plus, the seeded vessel would have almost certainly been lost to time by now by plate tectonics, the corrosive power of the sea, or the impact of hitting the planet. There's a number of problems inherent with the idea of our planet being 'seeded' for life, but putting them aside, the idea of us being able to prove it is basically nonexistent unless we find the remnants of some long lost civilization in our own travels/observations.

[@haskell6001]

​@@JO_______ No matter what anyone else says after me, you're keyed for giving an actual explanation.

[@globurim]

It would be a different language from us so it would take a while to even find it

[@i-use-4rch-btw]

Imagine if we all went extinct and aliens tried recovering our DNA to figure out what we were like but instead they accidentally found data storage DNA and became confused lol

[@troybarnes66]

And they'll learn what porn hub is. Once data storage using DNA chains is done you already know adult content is next.

[@Broken_robot1986]

I think it's like Prometheus.

[@johnnychang4233]

What if indeed the DNA already is an information system designed to 3D Print life as we know?

[@Nick-b7b9s]

Very hard to doubt Panspermia unless we are complete narcissists as a species

[@regu6582]

Could be the writing is already NOT on the wall, but in the DNA or RNA.

[@ToyotaKTM]

I think the ice cream went bad. NOO, you ate my hard drive!

[@stevengill1736]

It's OK, the DNA is in his body somewhere, just sequence a few tissue samples, etc. 😊 (actually, that's a great plot for a sci fi story!)

[@taiwanluthiers]

@@stevengill1736 Like Gattaca...

[@diegoantoniorosariopalomin2206]

Have you considered making a new video on alternative litography systems? Nanoimprint litograhy is finally shipping and and a few months ago a paper on far simpler Euv was published by Tsumoru Shintake

[@ristekostadinov2820]

i think he have talked about that Canon machine when it was announced

[@diegoantoniorosariopalomin2206]

@@ristekostadinov2820 that was a while back and it has since been shipped

[@ArawnOfAnnwn]

He should do one on Steady State Micro-Bunching too!

[@diegoantoniorosariopalomin2206]

@@ArawnOfAnnwn what is that?

[@itzhexen0]

All these people on the planet just shooting data all over the place like it's nothing.

[@MrCenturion13]

Just spraying g it everywhere....

[@publicspeaker4009]

Well in fairness I’m always leaving dna all over the place…. It gets everywhere

[@brodriguez11000]

Nine months later it gets backed up.

[@rkan2]

​@@brodriguez11000Better than tape 😅

[@justingrey6008]

There is a your mom joke in here somewhere

[@douro20]

Joe Davis gave a talk at the 30th Chaos Communication Congress where he talked about a project he was running at Harvard Medical School to encode part of Wikipedia into the genome of an an ancient apple species. He called the talk "Forbidden Fruit". He's a molecular biologist, btw.

[@densealloy]

Man attempts to store every season of South Park in DNA and inadvertently sequences Xenomorph DNA.😊

[@saxphile]

Great video as usual. I've worked with DNA synthesis and sequencing technologies for long enough to know better than assuming that the error problems won't be solved at some point. However, I believe DNA as a storage medium will never find utility outside of edge cases because too much computation will be required to encode and retrieve information (essentially zero for all other storage media), making latency and cost forever uncompetitive relative to contemporary technologies. Having said that, I think we should continue to work on it because it may work for a few edge cases where nothing else would be practical.

[@brodriguez11000]

Cold storage.

[@saxphile]

@@brodriguez11000 Maybe, but the conditions required to keep DNA stable for a long time would also work for many other storage media without the I/O overhead.

[@nischaymiglani2617]

Hey how your comment is 1 month old when this video is only 1 day old.

[@saxphile]

@@nischaymiglani2617 Patreon supporter

[@burntxela1258]

As a molecular tech, this is a very good summary of what the field is about, hope you cover more topics in the biotech/molecular tech alongside your silicon tech videos!

[@thunderdrumandbass]

Last December I got SUPER into this idea and even tracked down the HEAD of twist biosciences division that was responsible for this and she told me they sold off that portion!!!

[@jonahansen]

Why would you want to? Storage as well as retrieval is hard, slow, and expensive, and retention requires careful control of the environment. Using simpler polymers would be easier, but all chemical methods will need some way of interfacing to standard electrical systems if they are really being used for storage of arbitrary data. This would be easier if the monomers storing bits interacted with light in some easy, binary way as in compact disks, best engineered with special chemical forms that simplify synthesis. I think a lot of these studies are mainly for fun and to elevate the careers of the investigators. Nice video, though on what is currently being done.

[@CM-mo7mv]

full ack

[@idzkk]

Inventions do happen way before a use case

[@obsidianjane4413]

@@idzkk A lot of inventions have no use cases. The US Patent Office is full of them.

[@ashutoshsethi6150]

​@@obsidianjane4413 we want libraries

[@Zigglegarf]

Every time I consider the idea, it sounds like the worst tape drive imaginable. His quip "write once, read never" really sums it up. To me, it seems that the primary benefit of this research isn't data storage, or anything of the kind, but in helping to push the envelope on the sequencing, manipulation and production of increasingly complex artificial DNA.

[@thebeatnumber]

With that much storage power, you could effectively store evrery movie, every episode of every show or serial, every sporting event, every book ever written and every song ever recorded, on a device small enough to fit in the palm of your hand.

[@MrCenturion13]

Now to find the time to watch it all...

[@minus3dbintheteens60]

who is going to process all that?

[@obsidianjane4413]

And it only takes about a week to find and retrieve any specific file from it.

[@astrayelmgod]

@@obsidianjane4413 And a LARGE building full of very expensive equipment and highly trained and paid specialists.

[@testboga5991]

Biologist here: It's all investor hype because the stability is not high enough on the macro level. At the effort you have to put in, you can get better stability and access with different technologies.

[@bretthagey7916]

I remember watching someone at MS Research say some day you'll increase computer storage by growing another plant in a pot, and plugging in a wire. They are probably working on some wild DNA things now.

[@kellymoses8566]

MS research is currently working on storing data on glass slides for long term archiving.

[@dianapennepacker6854]

We need biological parts damn it. Imagine having bio machinery that constantly regenerated. Someone needs to design biological gears and ball bearings!

[@vilian9185]

Finally can install the new COD

[@johnnychang4233]

It's the plot device of an episode of Star Trek TNG titled 'The Drumhead' in which a Klingon spy used protein to encode and smuggle stolen information back to the Romulans.

[@tomholroyd7519]

Yup. Not to mention Voyager's biogel packs, because nobody knows what they are or what they do, except they got sick once, a disease called "plot device"

[@chrisbalfour466]

Every high power radio transmission gets stored in space, traveling at the speed of light away from Earth. But, realistically because of the inverse square law the signals are unrecoverable after a few years. I haven't heard anyone else suggest it, but three or more relays in space passing information in an endless cycle could store huge amounts of data. The distance between them determines how much they'd store.

[@WEPayne]

Synchronicity! Just last week started a Google sheet to study this very subject. Keep up your Most Excellent work !!

[@GeDruchy]

I remember going to a talk in Seattle in 2017 that was discussing (apparently then active) cyber attack vector against DNA sequencing instruments in a prominant US hospital. The attack went as follows: - An attacker encodes malicious computer code into some format (the format depends on the target victim machine) - The attacker then generates sythetic DNA based on that encoded sequence - The attacker injects their agent with the synthetic DNA, who then goes to the hospital for a blood test - The DNA sequencing machine reads the synthetic DNA - In the process of reading the synthetic DNA, the machine decodes the malicious code and inadvertantly runs it - The attacker now has privileged access to hospital infrastructre.

[@rollinwithunclepete824]

As always, Jon. An interesting subject!

[@rosshoyt2030]

0:42 wouldn't you want to read it at least once lol

[@tracyrreed]

Yeah that left me scratching my head too.

[@JorgeLopez-qj8pu]

17 Zetabytes? Finally! I can store 1/2 of a modern game!!!

[@hyperbitcoinizationpod]

This is like re-inventing the aeroplane using feathers.

[@klauszinser]

Very well done. Already getting this question/subject is a good idea.

[@undivided_unified]

So, what you're saying is the MATRIX real, got it

[@flopasen]

always has been need Bitcoin to succeed

[@robertthallium6883]

Great. That means we're not going to give equivalent human rights to robots when they show up requesting to the United Nations while dressed in their underwear. They'll get pissed off because we disassembled their envoys, brood a little in their "Zero One" whatever colony in the Middle East, and kill off a lot of humanity. Later, the robots learn they can turn our bodies into batteries, and then that's when Neo is rescued and Keanu Reeves buys a motorcycle in every county he visits for the rest of his life.

[@WASDsweden]

"...encoded a movie into the genome of living bacteria..." a new dimension of rick-rolling unlocked.

[@p.0-npcg.248]

The stability of DNA can also be thought as a natural form of differential signaling that we use for wired data transfer

[@periapsis413]

I would LOVE it if you did more videos on the potential future mediums for computing- I’m working on a sci fi setting and I’m already tempted to yoink this DNA storage idea…

[@mike-q2f4f]

Taylor Swift's music can now literally go viral.

[@stancil83]

4:04 Aw shucks. And here I was thinking I would be my own storage device.

[@MePeterNicholls]

6:43 “what’s the growing out your arm!” “It’s my iPhoto library. I accidentally touched it with a scratched finger”

[@Amorphous19995]

Although it might be plausible, as a biochemical engineer myself, I'm very skeptical that DNA storage could make any sense with the cost and effort required.

[@GrowlingBearMedia]

Write once, read never storage,... Perfect for my holiday go-pro footage !

[@geneballay9590]

another grrrreat video (as usual). thank you for all the work and then sharing.

[@Brighton24601]

You down with ATGC? Ya, you know me (And casting Uracil as the fifth beatle made me chuckle)

[@tjpprojects7192]

I wonder if our genome could be simplified. Junk D.N.A. doesn't "really" exist for the most part, but I bet if we 100% understood all aspects of our D.N.A., a LOT of functions could be simplified. Imagine a fully functional human with only one "standard-size" chromosome. Cancer, viral infections, and other things might be near non-existant at that point.

[@RichardFraser-y9t]

You are making some serious assumptions, those junk sections are a kind of checksum

[@tjpprojects7192]

@@RichardFraser-y9t Well I already said, "junk D.N.A. doesn't really exist", it's a fact that our D.N.A. could be simplified. Evolutionary pressures don't select for efficiency, but rather for what just works. With this in mind, saying, "we could simplify our D.N.A. if we 100% understood it" is NOT an assumption, it's a fact. By how much we could simplify it, I don't know, but we 100% could.

[@brianfong5711]

Is the read/write head gonna be a multiple nanomachines.

[@jansenart0]

I wrote about this idea in my 2014 novel.

[@XouXin]

Storing memes on DNA is how we get the first zombie apocalypse.

[@1.4142]

I wonder if you could make a computer out using DNA logic gates and storage.

[@HellothisisRev]

This stuff reminds me of wet computer concepts. Very good video

[@VVerVVurm]

All your base are belong to us!! 😅😅 i got the reference,, yay me!!

[@gsQartman]

JeffK lives on

[@AdvantestInc]

DNA storage feels like science fiction becoming reality. The challenge of cost-effective synthesis is huge, but the possibilities for cold storage are endless!

[@tastyfrzz1]

I actually worked on this at Seagate for a while.

[@kabirkumar5815]

what were you guys doing?

[@tastyfrzz1]

@kabirkumar5815 working on something that won't be around for a thousand years. First, imagine that every bit has to be on the order of a picoliter in size or smaller. Has to be, because a petabyte is 8,000,000,000 bits. Any bigger and you'd need train tankers of the chemicals. Then it has to be mind blowingly fast, like millions of chemical reactions, and deposition in perfect patterns on a perfectly clean substrate per second. Then there is the matter of transporting and moving the particles The only way I could imagine it was either with a cyclotron or paramagnetism. I spoke with physicists on this and even made a small demo of a piece of pyrolytic carbon floating over alternating neodymium magnets. It works. In the device I was imagining dibits written on a spinning high coercivity disk like how our servo tracks used to be and one would drop particles of bismuth or pyrolitic carbon over them and it would then accelerate out and shoot through a hole in a vacuum, into a tube that would take them through laser beam chillers or electrostatic chargers and then guided through numerous clouds of charged particles of the chemicals to build up the sequences then deposit or embed them onto/into a substrate. This thing would need to also be very small. Like a current hard drive. Needless to say, without a ton of funding the project wasn't going very far very fast. And just for fun, one of the physicists I worked with used to work at the Los Alamos National Testing Labs. He still provides them with magnets and chemicals and over one weekend did a study for me with powdered pyroltic carbon. He convinced me that a small cyclotron (while easy to make) would be too hot for this process. His initials are B.L. He actually helped me with a few different projects as he actually had diamagnetically charged magnets and knew where I could get cheap, perfectly flat mirrors!

[@stancil83]

11:34 Yes the 'ick factor' cannot be understated.

[@Wunderbolts]

Instructions unclear, accidentally added Shrek to my dog’s genome.

[@tonycosta3302]

I like the recent videos on memory storage. I’d like to see a video on holographic memory. It seemed promising years ago, but now you never hear about it.

[@Azu512]

Finally we won't have problem with storing Call Of Duty anymore, thanks guys!

[@JameBlack]

Just listened some related lecture on Simons Foundation channel. Time is flat circle

[@flopasen]

year 2089: i have a 2048 exabyte hard drive made from yeast

[@micgalovic]

I got, I got, I got, I got Loyalty, got royalty inside my DNA

[@youngmasterzhi]

In Horizon Zero Dawn, the scientists used this method of genetic data storage to hold all the knowledge of humanity, so that someday their AI programs will educate a new generation of humanity to regrow the devastated Earth, while they attempt to hack in and shut off the killer robots roaming around the planet All of their efforts were in vain after the sole cause of the robot plague purged all of that data and killed all the other scientists via air vacuum.

[@matveyshishov]

Can you please some day maybe make a video on that project silica from microsoft? I'm really curious because they made interesting statements on longevity and durability

[@timotheegoulet1511]

I’ve been wondering when you would come out a video about using DNA for data storage.

[@recklessroges]

Next I'm off to watch the film Species (1995).

[@wack1305]

Have you done videos or made comments on your research process? It would be interesting to see what I could learn from it

[@hitmusicworldwide]

I guess we could have stored a lot more information on the golden record on Voyager 1 and 2 and saved a lot of weight as well.

[@frogz]

all your base are belong to us!

[@code4chaosmobile]

Love the puns and jokes!! You rule!

[@frankholub4673]

Becoming the spiders from Children of Time

[@ragibmorshed6464]

3:06 they knew what they were doing when they named it that xD

[@perguto]

Imaging downloading a virus and it infect and kills you 😅

[@TrangleC]

Haven't watched the video yet (lack of time at the moment), just reacting to the headline, but I recently saw a video that said DNA doesn't actually store that much data. The human genome apparently are just a few Megabytes, like 60 MB or something like that.

[@Gokuguy1243]

matrix got it wrong, humans will become hard drives for our machine overlords, not batteries

[@thebandofbastards4934]

The original plot of Matrix was that people would be used as processing units. But they thought that it would be too hard for the average person to understand, so they went with batteries instead.

[@michaelmoorrees3585]

5:38 - So if you implement error correction, no Hulk, nor Spiderman.

[@tanini_pimpson]

Great video!

[@ElectricalInsanity]

I'm so much more concerned about this technology than any other. If you have the ability to create arbitrary genomes from scratch, a bad actor could create a real virus from software. Right now only a nation or large corporation could do something like this, but if the technology becomes ubiquitous it could lower the barrier to entry low enough for a motivated enthusiast to be successful.

[@TheShorterboy]

yeah when I get 3.5inch a biolab in my PC until then probably never happeing

[@ccshello1]

Step 1, encode and synthesis the entire encyclopedia into a long DNA strain. Step 2, insert it into a bacteria for safe keeping. Step 3, one day later, The Bacteria emerges as a Superman-like being, taking over the humanity. ------The End------

[@KSayar]

This is one of those crazy ideas that sound good on the surface. With this method you can only make write-once-never-read memory. İt is write-once, because, once the polymer structure is fixed, you cannot change it. İt is never-read because there is no technology that can read even a short chromosome end-to-end. Existing DNA readers can at most read a few hundred nucleotides from a Billions-long DNA at a time. Error rate in reading is so high that the reader needs to read the same segment dozens of times before they can apply computer algorithms to guess the correct sequence that they have read. Reading and error-correcting one chromosome takes weeks to complete -hardly competitive with the current hard drive technology.

[@Toksyuryel]

I'm curious what the advantage of DNA data storage is over grey matter storage, since the latter is actually _meant_ for data storage and we know it can be very very dense.

[@Telencephelon]

Hell yeah. All the base are belong to us!

[@Cs13762]

this is strictly for data storage and not genetic modification the same way OpenAI is strictly for non-military non-surveillance purposes: The thing making it not be used innapropriately is the fact that they have decided not to at the time or, at least, decided to claim that they are not. If and when someone decides to use it innapropriately it will be a highly developed and efficient method of achieving things on either side of the moral spectrum. clearly, the inevitable future is to use actual organisms to biologically maintain and preserve data. There is really no limit in the imagination as to what else might also be acomplished with the knowledge gained in applying those techniques. If people think AI is the most important thing to change the world, think again. Transistors are poppycock.

[@klauszinser]

Why: write once (n)ever read ..?

[@Nope_handlesaretrash]

1:57 MOVE ZIG MOVE ZIG

[@kuronyaa-san]

Bioneural gel packs when?

[@vegasu9418]

"plenty of room at the bottom" brought back memories :,) professor of my mems course in my alma mater recommended reading it at the start of the semester. oh how times fly :,)

[@thirstyCactus]

Anyone watch "Travelers"? They used this idea as well.

[@hibbs1712]

We have to be very careful how closely we tread to eugentics.

[@rainbowhyena1354]

I wonder if it's possible to make bacteria with its own genetic randomizer. To speed up evolution by crossing the Hamming distance to a beneficial mutation. Similar to how it randomizes the section of T-cell genome to produce different antibodies

[@BobSpector-up7lw]

Thanks!

[@beautifulsmall]

Is storing data in a structure reducing enthalpy so at some point energy could balance . a gravitational structure that was stable. like a bucky ball.

[@biosurveillance]

Maybe an ftl comms video?

[@alexis1156]

How fast can the data be red though? In a way that can be compared to current drives?

[@yourfriend4104]

I'm about to become a server drive 😭

[@ImKinoNichtSabbeln]

I wouldn't exactly freak out over a failure rate of "just" 0,5 % per bit. Apart from enforcing exorbitant huge overhead for redundancy, it also explodes the costs for writing, retrieving and processing the raw data exponentionally. After all, it's wet, organic chemistry to be artificially controlled at single-molecule level, and controlling literally every single side reaction at single-molecule level, too. And that included not even looking at the additional bloated overhead to avoid instable, and more importantly biological activite sequences, again just hinting at the fact that we don't understand more than a few details of the real biological interdependencies of DNAs and RNAs in the complete wild (not in models) in the first place. Remember, 40 years ago we did not know that bacteria freely exchange resistancy genes between unrelated stems. Dreamers of DNA based information storage might want to look back: 30 years ago, we didn't understand the function of the "inactive" part of the DNA, and classified 90% of human DNA as "garbage" that could easily could have been left out. The overlooking of the essential function of the microbiome, as well as meta-genetucs is also a testament to be by far more humbleness. I.e., if we accidently fuck up DNA because we still don't know/overlook/ignore something, then it's game over for the entire life on planet Earth. Oh, we know so much regarding DNA, RNA, and stuff? How come that we don't understand more than at best the bare minimum at single molecule level of the at least few hundred side reaction in this messy soup called cell when we struggle to understand the bio-molecular mechanics of but all diseases in the context of all biochemical reaction? If everything was so clear and well-known, how come there are still clinicar trials, that e.g. boner pills we accidently found, and there are so many unhealable diseases? That CRISPR-CAS is that new, and we are so often caught pants-down by biology? As humans we are completely incapable of anticipating any results of messing with complex systems. And Earth's biology is by far the most conplex systems. The whole blind enthusiasm reminds me of the nuclear enthusiasm of the 1950ies, but on fentanyl times a billion. And the hollow, absolutely one-eyed promises are the same as for dreamong of "usable" black hole to produce energy with maximum efficiency (46.5%, afair). Either way, it's super-criminally negligent. Conpared to this, childtem "baking" cake from sand is more related to TSMCs operation. Nearly 40 years ago, I studied chemistry. Back then, DNA information storage was the 2010th/20ies big think in data storage. And everybody was even imagined to carry around the personal data of a few 10 GB (tops!) is their own genonome by repurposing the "garbage" DNA. That aged well! Don't get blindfolded by models! Models are intentionally constructed to leave out hindering stuff to prove a principle/isolated detail. Whereas complete calculations in chemistry are a nightmare (thus we always, _always_ simplify massively, excepty for very small molecules), complete biochemical systems are a complete different universe of complexity. You shouldn't believe, that e.g. all the colourfull screw-pictures are exactly calculated. Nope. Just good, but still - in their core - happy-sloppy estimates (see above: Surprises and unsolved mysteries). Remember; Iit's just models. If there's a "screw" we don't know/forgot/rendered irrelevant, it could easily spoil everything. And biochemistry is littered with "screws" (yes, it resembles game theory's Dark Forest).

[@alekseyburrovets4747]

Have you actually read the book on molecular biology? I wouldn't believe you did.

[@JoseLopez-hp5oo]

My DNA is mostly NOP instructions.

[@MatthewSuffidy]

Probably more complicated than it is worth. You know using this exact tech though, say you had the COV19 genome and just transcribed it as an RNA strand. Then you injected the RNA strand using something like a needle into a cell. You would then have started COV19.

[@SteepSix]

I think using DNA to store data will prove to be impractical. Using the structure of DNA as a framework to develop new data storage techniques is more likely. After all, data storage in just 30 years has progressed beyond all imagination... What then cost many thousands of dollars and occupied enough space to fit a full-size fridge now fits in a wafer that's smaller than my thumb nail and costs less than a meal in an average cafe! What exactly do we need this exotic DNA technology for?

[@the_primal_instinct]

You dawg, we heard you like viruses, so we put a virus into a virus.

[@buckmazz]

Moveiplot: "Scientist wamts to create fancy harddrive... summons "Venom" by accident..!"

[@1.4142]

Could we engineer our own nucleotides that are more stable and less error prone?

[@JohnnyWednesday]

What if Earth is some alien species hard drive?

[@iankirby4160]

I can imagine a meme being encoded and turn out to code for a prion disease. Literal brainrot.

[@将軍九八.彁]

Could bits of random data create viruses?

[@LukasSmith827]

The matrix... IS REAL?!

[@joaopereira685]

Just a what if, but if this DNA somehow gets into our bodies, like say through a virus, could this not mean perhaps something disastrous? If it were the case that all the DNA in the world has some basic necessary structure that whatever we store in the DNA doesn't have there could be really unpredictable consequences.

[@catcatcatcatcatcatcatcatcatca]

Or we could use any other polymer as well. DNA might not be the best canditate, especially if we don’t care about other properties of it such as replication.

[@bencordell1965]

That was my idea