Skip to main content


Better, Faster, More Comprehensive Manure Distribution

So today brings news that Microsoft is working on curing cancer in the next five or ten years. That, I’m sure, will come as a relief, especially to those people who’ve had the company’s software crash on them recently. For some reason, the UK press is especially susceptible to Amazing Cancer Cure stories (and to Amazing Alzheimer’s Cure stories, too, for that matter) so you get coverage like this in the relatively sober pages of the Telegraph:

Microsoft has vowed to “solve the problem of cancer” within a decade by using ground-breaking computer science to crack the code of diseased cells so they can be reprogrammed back to a healthy state.

In a dramatic change of direction for the technology giant, the company has assembled a “small army” of the world’s best biologists, programmers and engineers who are tackling cancer as if it were a bug in a computer system.

. . .The programming principles and tools group has already developed software that mimics the healthy behavior of a cell, so that it can be compared to that of a diseased cell, to work out where the problem occurred and how it can be fixed.”

Have they now? It’s hard to tell from the article how much of this is hype from Microsoft and how much was added in the reporting. So I turned to a story at Wired UK, which features many of the same word-for-word quotes as the Telegraph piece, and saw this:

Andrew Phillips, head of the biological computation research group at the Cambridge Lab said researchers benefit from Microsoft’s history as a software innovator.

“We can use methods that we’ve developed for programming computers to program biology, and then unlock even more applications and even better treatments,” he said.

Phillips is working to create a molecular computer that could be put inside a cell to monitor for disease. If the sensor detected a disease, such as cancer, it would activate a response to fight it.

Right. What about something on this side of the Atlantic? Here’s CNN Money, and I don’t think the promises are getting dialed down much:

Researchers at Microsoft’s lab in Cambridge, England, are trying to map the “code” of the disease, hoping that once they understand how the problem occurs, they’d be able to re-program cancer cells into healthy cells.

Jasmin Fisher, who is a senior researcher at the lab and also a professor at the biochemistry department at Cambridge University, said she and her colleagues are trying to think about cancer in the same way computer scientists think about computer programs.

“If you can figure out how to build these programs, and then you can debug them, it’s a solved problem,” she said.

Well, I can’t argue with that last part – if you can figure out what’s going wrong, and then figure out how to fix it, then you have indeed solved the problem. Where I part company with the Microsoft press coverage is the idea that it’s going to be as straightforward as writing and debugging code. And yes, I realize that writing good code (and chasing down the inevitable bugs in it) is actually hard work. But guess what, folks? Trying to figure out cell biology is even harder.

I have beaten on this theme many times on the blog, so for those who haven’t heard me rant on the subject, let me refer you to this post and the links in it. Put shortly – and these sorts of stories tend to put actual oncology researchers in a pretty short mood – the cell/computer analogy is too facile to be useful. And that goes, with chocolate sprinkles on it, for all the subsidiary analogies, such as DNA/source code, disease/bug, etc. One one level, these things do sort of fit, but it’s not a level that you can get much use out of. DNA is much, much messier than any usable code ever written, and it’s messier on several different levels and in a lot of different ways. These (which include the complications of transcriptional regulation, post-transcriptional modification, epigenetic factors, repair mechanisms and mutation rates, and much, much, more), have no good analogies (especially when taken together) in coding. And these DNA-level concerns are only the beginning! That’s where you start working on an actual therapy; that’s what we call “Target ID”, and it’s way, way back in the process of finding a drug. So many complications await you after that – you can easily spend your entire working life on them, and many of us have.

And that’s why many of us who have actually been working on diseases like cancer get a little testy when we see folks from computer science coming in with this “Gosh darn it fellows, do I have to do everything myself?” attitude. Years of working with (human-designed) hardware running (human-written) code have given many people in that field what I think is an exaggerated idea of human capabilities (at least as they are right now). When you can write code that gets used by hundreds of millions of people in their daily lives, it’s understandable to think that you’re able to just reach in and change reality by sheer braininess and force of will. Unfortunately the world of code and computational hardware, as important, useful, and lucrative as it is, is just a sandbox compared to the real physical universe, of which living creatures are just a tiny little part. But biology has no debugging programs, no annotations, no manuals. It wasn’t written by humans – in fact, as far as we know, it wasn’t written by anyone at all, it “just grew” in a process that has no good counterpart to the ways that humans generally get things done.

I like borrowing examples from old science fiction stories, and William Gibson’s early work is, I suppose, heading into that category by now. Here are a couple of characters (well, one human one and one recorded human personality) in Neuromancer discussing their dealings with a (malevolent?) artificial intelligence:

“Motive,” the construct said. “Real motive problem, with an AI. Not human, see?”

“Well, yeah, obviously.”

“Nope. I mean, it’s not human. And you can’t get a handle on it. Me, I’m not human either, but I respond like one. See?”

“Wait a sec,” Case said. “Are you sentient, or not?”

“Well, it feels like I am, kid, but I’m really just a bunch of ROM. It’s one of them, ah, philosophical questions, I guess….” The ugly laughter sensation rattled down Case’s spine. “But I ain’t likely to write you no poem, if you follow me. Your AI, it just might. But it ain’t no way human.”

Exactly. The only quote by the Roman poet Terence that I have at the ready is the one that translates as “I am human; nothing human is alien to me”. Well, this may sound odd to the group at Microsoft, but cell biology is alien to you. It’s alien to all of us. We’re not even near understanding what’s going on in normal cells or cancerous ones, so giving people the impression that you’ve already simulated everything important and you’re busy “debugging” it is not only arrogant, it’s close to irresponsible. It’s like those stories about simulating the brain – sure, dude.

If you remove the hubris from the Microsoft announcement, though, which takes sandblasters and water cannons, you get to something that could be interesting. It’s another machine learning approach to biology, from what I can make out, and I’m not opposed in principle to that sort of thing at all. It has to be approached with caution, though, because any application of machine learning to the biology literature has to take into account that a good percentage of that literature is crap, and that negative results (which have great value for these systems) are grievously underrepresented in it as well. I think that machine approaches to understanding biological pathways will, in the end, probably be the way to go, because it’s too complex for us to keep it all straight in our minds (not human!) But we’re not there. There are many, many important things that we simply don’t understand very well, and many others, I’m sure, that we just flat-out don’t even know exist yet. Debug that.

So if Microsoft wants to apply machine learning to cancer biology, I’m all for it. But they should just go and try it and report back when something interesting comes out of it, rather than beginning by making a big noise in the newspapers. You want to cure cancer? Go do it; don’t sit around giving interviews about how you’re going to cure cancer real soon now. I’m sure that someone at the company imagines this as a big blazing publicity sendoff, fireworks and balloons and all the rest of it. But to anyone who works in the field, it’s more like a grenade going off inside a manure pile. Or this. Probably not what they were aiming for.

Update: one of several relevant XKCDs!

120 comments on “Better, Faster, More Comprehensive Manure Distribution”

  1. Anonym says:

    I was waiting my whole work day for your take on this (already 3 p.m. here in Germany), so thanks for delivering and making my day! (Well, for the second time, the Telegraph article in the morning already caused a lot of amusement)

  2. Curious Wavefunction says:

    Evolution makes mistakes, and these mistakes are responsible for the amazing functions that are encoded into DNA. Most of life’s software is messy, multilayered and buggy because it’s had 4 billion years to evolve this way by fits and starts. Perhaps the best way for a machine learning algorithm to try to replicate DNA’s functions would be for its designers to allow it to make mistakes and incorporate their results as fast as possible.

    1. loupgarous says:

      Microsoft is failing to realize that by treating DNA as buggy code, they’re literally dicing with death. Each and every gene and codon in our genome was selected naturally, which means that the genes and codons that didn’t make the cut are in misshapen cousins of ours out in the Olduvai Gorge, waiting for someone to dig them out.

      Death is the natural selector, and if you’re lucky and your genes don’t make the cut, you just don’t transmit them to future generations (genetic death). If you’re less lucky, you get cancer or some other mutation incompatible with life, and your genes go away with the rest of you in a pine box (lead-lined oak, if Mom and Dad can afford it).

      Microsoft is blithely proposing to transfer their model of software creation to cell biology. Someone beat me to “this gives ‘the blue screen of death’ a whole new meaning,” and that person was right. When Microsoft does what they say they will, recoding cancer cells so they’re normal… what happens when we get the cellular equivalent of the first Windows Vista release? Death.

      Microsoft ought to concentrate on debugging their existing software before they branch out into wetware. I notice in the toolbar of Windows 10 on my laptop that the clock in the lower right corner just stops keeping accurate time now and then. I think of the equivalent malfunction in, say, the ribosomes in my cells – and cringe.

      1. Nick K says:

        If Microsoft can’t even debug Windows 10 properly, what earthly hope do they have with the infinitely more complex and re-entrant code of living cells?

  3. Billy says:

    Didn’t Verily already do this? They should be about 80% of the way to cure by now. Oh wait, that small army of the “best biologists” left in the last year so maybe it is in the hands of Microsoft. smh

  4. Dave says:

    That will give the phrase “Blue Screen Of Death” a whole new meaning.


  5. Hap says:

    1) Shouldn’t the phrase “domain of applicability” have come up somewhere in these discussions? Using logic where it doesn’t apply (or where you don’t know what logic applies) seems like a recipe for failure. And limitless arrogance.

    2) People get annoyed when your software doesn’t work and you don’t care, but people mostly live with it because they have to and because it helps them do things that they could not do easily (or at all) otherwise. Health care has a different level of fault (in)tolerance.

  6. SirWired says:

    Proving the adage that there’s an appropriate xkcd comic for everything:

    1. SirWired says:

      Heh. And because the xkcd guy is a mind reader, TODAY’s comic might as well have been written specifically for this article!

      1. Anon says:

        I, too, was just about to post that same link!

        Pity MS haven’t seen it.

      2. Curious Wavefunction says:

        I guess what that xkcd comic is saying is that you can kill cancer cells using an atomic bomb. Makes sense.

        1. SirWired says:

          Read the “alt” text. (Just leave your mouse cursor parked over the comic.)

          (xkcd always has a second punchline in the alt text)

        2. Barry says:

          killing cancer cells is easy. Discriminating cancer cells from host cells is tricky.

    2. Anon says:

      That one is brillian, and spot on here!

  7. loupgarous says:

    First, Biden’s cancer moon shot (just one vowel wrong in that last term), now Microsoft’s betraying their massive ignorance of oncology with this garbage.

    Note to Microsoft: Want to know how dumb you look? Imagine a bunch of oncologists – the leaders in the field – announce with great fanfare they’re going to use their Petri dishes and Western blots to make a world-class operating system for laptops that blows Windows 10 away.

    That’s how dumb you look.

    1. oldnuke says:

      Actually, it’d bet on the oncologists as being more likely to succeed than MS.

      1. loupgarous says:

        You’re right. Oncologists almost all know to use SAS, Oracle, or another well-validated platform to do statistical testing. Not Microsoft Excel.

    2. Nick K says:

      Nail on head. What amazing arrogance on the part of Microsoft.

    3. Dwywit says:

      Those of us in general IT work are used to the shenanigans of software companies, and Microsoft is a leading light. It happens with software all the time – fanfare, fireworks and a f*ing big announcement that they’re going to release this new, magnificent, re-built-from-the-ground-up piece of razzle-dazzle that will just…… change your life, man!

      And then 6 months down the track comes the next version of MS Office, or Windows 10, and we all breathe a sigh of disappointment.

      I was going to say – “If they want to throw their billions at cancer research, great. It can’t do any harm, and it might help”, but then I realised that it *can* do harm, and great harm, by affecting through sheer momentum the course of highly valuable traditional research.

  8. Crick says:

    Unfortunate, Watson was right again:
    “most of the experiments we do are irrelevant … We’re not going to cure cancer by doubling the money. We’re going to do it by being more intelligent. The money thing is just a red herring of people not thinking.”

    1. Anon says:

      Hope you don’t mean IBM’s Watson – because that thing is as thick as sh*t.

      I tried asking it “who won the Brexit vote?” and “what was the UK referendum result?”, and instead of just answering the actual question, it did all sorts of useless stats on most active accounts tweeting about the referendum – including all those before the vote was made.

      Meanwhile a quick news search on Google gave me the answer in 2 seconds.

      If I interviewed Watson for a real job, it would be thrown out the room even faster.

      1. Crick says:


  9. dearieme says:


  10. oldnuke says:

    Well, hopefully they will try their “cure” on themselves first. That way we’ll solve two problems in one sitting.

    They need to stop inhaling whatever they are smoking.

  11. experimentalist says:

    I hope other tech companies follow. Maybe the folks at Tindr and Grindr team up to lend the raw passion of their clients for curing cancer?

  12. Kent G. Budge says:

    ““If you can figure out how to build these programs, and then you can debug them, it’s a solved problem,” she said.”

    I make a comfortable living basically by debugging code. Oh, sure, I write it first — but the hard part is finding and fixing the bugs. It’s a real art, which is amazingly difficult to do, and requires both a certain knack and lots of experience. Which is why the comfortable living.

    And my code pretty much exists to solve a single well-understood physics equation and some of its variations and approximations.

    So my reaction is, “Oh, is that all it takes?” in the most sarcastic voice possible.

    1. Charlie Kilian says:

      This is exactly what I was coming here to say. I write code and debug it for a living. I know a lot about this topic. And I’m baffled as to what they think they’re saying when they say “we know how to debug code.”

      Do we? Because that actually isn’t true, at least not in any way that translates to biology. If we know how to debug code, it’s mostly because we have gotten better at writing code in the first place. What we’ve come to know how to do is avoid the really hard bugs. We know how to write code to minimize the chance of bugs happening. We know how to avoid complicated code that leads to the worst kinds of bugs. We know it’s a bad idea to write self-modifying code, for example, and we also know that code that tries to do two things at once and interact with itself (anything with multiple threads sharing resources) complicates the picture enough that it’s always a good idea to think long and hard about whether there is a different way to write it. To whatever extent possible, we avoid interconnectedness, and strive to make each piece of code do only one thing. We also write digital code, and have almost no experience (relatively speaking) writing or debugging analog code, which biology undoubtedly is. None of our expertise in writing and debugging code applies to biology, which even if I grant the computer programming analogy, is code we didn’t write using methods we’ve never seen. It does all of these bad things we’ve learned to avoid in our own code and more. It does things so bad, so convoluted and interconnected, we’ve never even thought to do them.

      All of which is to say, these people don’t strike me as particularly clear thinkers even when it comes to their own problem domain.

      1. John Wayne says:

        Most pharma companies are run by lawyers and MBAs with no science background; I’m sure that the folks that run Microsoft are similarly disconnected from their core business. This is one of the results.

        Thank you very much for leaving comments from the perspective of a coder, it is really interesting that self-modifying code and code with interdependencies are things to be avoided. In biology, those things are the rule.

      2. HFM says:

        I was trained as a synthetic biologist, and dealt with a lot of engineers trying to port engineering tools to biology. I think that “intelligent design” (for lack of a better term) is such a massive design constraint – at some point, a human-designed system made sense to a human being somewhere, and there were less-sensical alternatives that weren’t pursued. Biology just has to work; it’s a whole different beast. But people who’ve spent their professional lives dealing exclusively with human-designed systems have a hard time seeing it; they’re so accustomed to design, it’s like explaining water to a fish.

  13. Bagnar says:

    This is an excellent idea ! Cancer research will be useless. Researchers will have to recycle themselves in… Why don’t they try to debug any Microsoft software for example ? There is a plenty of work in this area !

  14. myma says:

    I think we would all do just as well as MS will if we just had one of those 12hp 4500watt Noise and Fume generators from “Harbor Freight”.

  15. DCRogers says:

    The Microsoft source material for those news articles can be found here:

  16. Jason Kahn says:

    The little computer sitting in a cell waiting to fix it was a fixture in Drexler’s Engines of Creation. Maybe the Microsoft folks read that as a textbook by mistake.

  17. enl says:

    Surprisingly enough, When I opened this entry this morning, Windows decided that it was time to crash….

  18. asoc says:

    The best quote was from the Telegraph story:

    “I think for some of the cancers five years, but definitely within a decade. Then we will probably have a century free of cancer.”

    1. Nick K says:

      The scientific illiteracy of British newspapers is astounding, as is their credulity.

      1. John Dallman says:

        Their business model is not based on truth. In the UK, TV has duties about fair reporting and newspapers do not. I gather in the US it’s the other way around?

        The Telegraph, Mail and Express have readers who are mostly old, and thus assumed to be afraid of death, and who don’t seem to remember much. So stories like this seem to reassure them that things will be fixed and they won’t get cancer. Yes, there’s no logic at all there. Welcome to the UK press.

        See also the Daily Mail ontological oncology project:

  19. Isidore says:

    This reminds me of the old joke, ca. late 1990s, but applicable today as well, that if the development of automobiles had paralleled that of computers, cars would cost $1000, they’d get 200 miles to a gallon and at random intervals they’d blow up killing all passengers.

    1. myma says:

      Reminds me of another old joke:
      There are 4 engineers in a car. Car won’t start. The electrical engineer says “I know, its the battery/alternator, we need a jump start”. The chemical engineer says “I know, its the fuel pump, let’s fix that”. The mechanical engineer says “I know, its (… something mechanical)”. The computer engineer says “I know, I know, how about we all get out of the car and back in again and see if that works”.

  20. neo says:

    I’m curious. The first time I ever read “we will cure cancer in 10 years” from a credible source was 1972. Can anyone beat that?

    1. Kent G. Budge says:

      Not, not in this field.

      But in my own field, controlled fusion has been fifty years away since around 1952.

      1. loupgarous says:

        And $80 billion later, big-iron controlled thermonuclear fusion’s only a heartbeat away. Just like it was in 1952. Robert Bussard once admitted his work on controlled fusion was basically a way to do more fascinating science, not to make Reddy Kilowatt more ready.

  21. Bill Gates says:

    Is it just me, or is complete fantasy in fashion with investors right now?

  22. David Cockburn says:

    It would be sad if it wasn’t laughable.

  23. Earl Boebert says:

    From the computer geek site Slashdot, where Microsoft’s PR is getting the treatment it deserves:

    “Have you tried dying and being reincarnated?”

    — Tech Support

    1. Nick K says:

      Thank you, that has made my day!

    2. loupgarous says:

      Tech Support Guy at Microsoft’s Blindingly Obvious Cures for Cancer Division:
      “Hi, I’m Chip. Ok, you’re a treatment failure in non-Hodgkin’s lymphoma. Hang on, let me pull that up… I’m going to walk you through a cure… oh. Can you hold, please?”

      (two months later…. ) “Hello? Hello? Are you Patient Jane Doe?”

      Response: “This is her mom. Can you call back later? We’re having her wake.”

      Tech Support Guy: “Oh, good. I was going to suggest she reboot… “

  24. Emjeff says:

    It’s an old problem, probably as old as humanity itself. Because you’ve had some success in one field ( science, music, acting), you then think you are qualified to be as successful in other fields. This is why we have actors opining on vaccines and nutrition, and physicists insisting that Vitamin C can cure everything. It is the highest form of hubris there is, and it does a lot of damage.

    1. Anon says:

      As Stephen Hawking said: “The greatest enemy of knowledge is not ignorance; it is the illusion of knowledge”. But then WTF would he know!!

      1. Hap says:

        I thought that was Daniel Boorstin (unless there’s some joke here that I’m not getting).

  25. Eric says:

    This reminds me of my son’s two step process for solving any problem:

    1) admit you have a problem
    2) fix the problem

    I’m thinking my son should found a start-up to cure cancer.

    1. Derek Lowe says:

      I’m going to regret making this comment, but that sounds like Donald Trump’s plans for almost every problem he’s spoken about. He is, apparently, going to fix Problem X by fixing Problem X, with something “really great”. Next problem?

      1. problem solved says:

        Derek that’s it …. Microsoft will build a wall around cancer and the Donald will pay for it!

        p.s. thanks for the posts

        1. NJBiologist says:

          That’s it–*encapsulate* those gliomas! Why didn’t any of those fancy oncologists think of that before?

          Sorry, I had to.

      2. loupgarous says:

        Unfortunately, it’s a bipartisan syndrome. Joe Biden called for a “moon shot” against cancer, which shows he wasn’t paying attention to what a dead end Project Apollo was in practical terms. Richard Nixon and his scientific advisers looked at the numbers, decided we couldn’t colonize the Moon, have a defense program that deterred our enemies, and fund the obligations created by the “Great Society”.

        They never got due credit for making the Great Society work as well as it ever would, while building the space transportation system the country needed (our dominance in military space with the help of heavy spacelift from the Shuttle probably helped win the Cold War). Historians may look back at our current politics and thank the plodders, not the showmen.

        “Moon Shots” and “Yuge Walls” look good in the news and waste money. Methodical, unglamorous, carefully-paced research and development has given us every great advance against cancer we now have (and I’m sincerely grateful for that).

        1. Design Monkey says:

          Project Apollo had only one practical goal – to show off who really has a longer dick. (was kind of necessary after initial soviet gains in the same grounds) . It was perfectly successful. Soviets wimped out both technologically and economically.

          1. Duane Schulthess says:

            But at least we got Tang for a few years… wonderful powdered orange drink flavored with tin.

            Perhaps Microsoft could daisy chain together the millions of left over Zune’s they have laying around for the computational power for their cancer effort?

  26. vasantha says:

    It’s quite simple. Silicon Valley made Blood—Simpler. Now they want to make Cancer—-Simpler

  27. Dr. Manhatten says:

    CD/ C:
    C:/ >Run cancer cure
    C:/> Bad command or file name

    1. John Wayne says:

      Everybody who laughed at this comment is old (including me)

  28. maw says:

    That picture in the telegraph shows the PipetMax. Possibly the worst liquid handler on the market.

    Maybe MS should pony up some money for people and equipment instead of generating PR.

  29. Cato says:

    So basically we can expect this in ten years right?

    credit: Elysium

    1. Anon says:

      Yep, including the fact that only 0.000001% of the population would be able to afford it.

  30. Chrispy says:

    Microsoft has not even managed to get right the very things they have had all the tools to get right. All they have ever managed to do is imitate and bully. Companies like Google and Apple make amazing advances that blow right past Microsoft, which gives us Bing and the Windows phone. Meanwhile, you can’t even do simple things like print an excel spreadsheet without formatting errors.

    Microsoft: get your own damn house in order before proposing to work on ours!

  31. Rob says:

    I think a good way to talk to computer scientists on the complexity of biological systems is to provide the following example of evolutionarily-produced programs on FPGA chips. They can program themselves after 4000 generations, but no one could really figure out how they worked (from the article linked):
    “It seems that evolution had not merely selected the best code for the task, it had also advocated those programs which took advantage of the electromagnetic quirks of that specific microchip environment. The five separate logic cells were clearly crucial to the chip’s operation, but they were interacting with the main circuitry through some unorthodox method— most likely via the subtle magnetic fields that are created when electrons flow through circuitry, an effect known as magnetic flux. There was also evidence that the circuit was not relying solely on the transistors’ absolute ON and OFF positions like a typical chip; it was capitalizing upon analogue shades of gray along with the digital black and white.”

    1. MTK says:

      I like that Rob and was thinking along the same lines but in more simplistic terms.

      Essentially, imagine a vintage 1980’s IBM PC that is still able to run just as well as any current laptop through countless hardware and software changes, recasing, and other modifications and upgrades and that these upgrades have all occurred as a response to competition and the environment with no planning or human intervention at all. Essentially, these have all been patches of some sort. Try fixing or debugging that if something should now go wrong without somehow screwing up some other function.

      Except a biological system has billions of years of accumulated patches, artifacts, and redundancies. Nature, for the most part, has not reformatted its hard drive or installed a new operating system in order to start from scratch. It just continually builds or modifies existing hardware and software.

      1. Dr. CNS says:

        … such a complex system…
        Isn’t it amazing that a daily dose of 10 mg of a white powder could treat these terrible diseases…

        Or can it?

        1. Anon says:

          At least those 10 mg therapeutics are giving hope to stage 1 – III patients of extended life 1-10 years!

          If these billionaires really want to have some corporate / social responsibility, they should just fund the funding agencies to fund the basic and applied researchers to enble them do best they can than insulting them by reducing everything to the “coding…debugging etc”. It is comical at best!!

          1. Bob says:

            Yes. Just ask Microsoft to pay their taxes, The NIH will take care of the rest.

    2. Mol Biologist says:

      IMO To make it more biological relevant I would consider analogue shades as parallel genes transfer and the possibility of evolutionary breakthrough. So new circuits or biosynthesis pathways can be created and some will be disrupted and eliminated. My question is what circus will be kept and I think probably one due to survival advantage. I am kind of confused with the statement that “It seems that evolution had not merely selected the best code for the task”.

  32. Calvin says:

    Ctrl Alt Delete

    Must be that simple. Surely.

  33. Ech says:

    If DNA is source code, then it was written, nothe by the typical design process used at MSFT, but via genetic programming. It’seems messy, full of dead ends, poorly organized, and has no comments.

    I did a research project at a big US space agency to use genetic programming techniques to evolve control programs for a robotic hand. What we ended up with were programs like what I described above.

  34. Clayton says:

    Honestly my biggest hope in life is that we can someday name a suitable asteroid “Ctrl Alt Delete” for the appropriate reason. Even if it only destroys the West Coast of North America.

  35. Rule (of 5) Breaker says:

    I have worked in cancer research for almost 20 years, and it turns out all I had to do to cure cancer was enlist the help of an anthropomorphic paperclip.

    1. loupgarous says:

      To be fair, the talking paperclip and SCSI the Rat were stuff Gates let his Mrs-to-be develop, weren’t they?

      Which raises the question “Whose romantic interest developed this idea?”

  36. Canman says:

    Be honest, most people enjoy watching ‘big’ people fail. Just sit back and carry on doing what you’re doing and wait for MS’s big fail in a few years. There’s no sense worrying about it, and we all know it’s inevitable.

  37. Andrew says:

    One thing you can reliably count on Gibson for is not actually knowing the first thing about computers.

    1. Derek Lowe says:

      Hah! Neuromancer also features an orbiting resort with pay phones in it, among other oddities. I like it anyway!

      1. Diver dude says:

        The pay phones in question were in the lobby of the Istanbul Hilton.

        My. Favorite. Book. Ever.

        1. Derek Lowe says:

          Right you are! I wonder if they still have any?

  38. Anon says:

    Why stop at cancer? Pathetic! Zuckerberg’s going to cure … Everything!

    1. Curious Wavefunction says:

      To be fair, his exact words were:

      “Can we help scientists to cure, prevent or manage all diseases within our children’s lifetime? I’m optimistic we can.”

      Still optimistic (especially the timelines), but using the words ‘prevent’ and especially ‘manage’ tempers the enthusiasm.

  39. DrSAR says:

    I’m just dismayed to see how low they’re aiming. Cancer? And that’s all? Look to someone like Zuckerberg – he is going to just deal with disease. All of them. Gone by the end of the century.
    Apparently $3E9 are going to solve it which I though is currently the price tag for 1.5 drugs. So this must be some super pill he is planning…

    1. loupgarous says:

      Mr. Zuckerberg’s short on sales resistance. He also peeled off $1bn for Oculus, a 3-D VR gear manufacturer, in a market space full of people like HTC who are used to making expensive things like cell phones cheap. The only product to see the light of day from Oculus is their “Rift” 3-D VR system. So far, the virtual world Second Life and another virtual world with a seamier customer base are the only two places Oculus Rift buyers will have a custom-made experience – when the two virtual worlds beat the bugs out of the “port” of their PC-only current versions to a more VR-friendly format,

      Which shows the problems Microsoft and Zuckerberg don’t seem to see ahead of them – if a session in Windows or Second Life crashes, it’s annoying – in the case of Windows, you might lose work, or if you depend on Excel to do what you should be doing in SAS or Oracle, you could look truly stupid before other professionals.

      If a drug these people design using a pick-up drug development team somehow squeaks past tox and the study team isn’t absolutely rigorous in animal, healthy volunteer and patient safety/efficacy testing. they could get into the brave new world of not just annoying people, but killing or crippling them. I wonder how long it’ll take for these guys to get the sang-froid you see in some of the current players in Big Pharma.

  40. Imaging guy says:

    Even if we cured all cancers-“Likewise, elimination of heart disease would increase
    life expectancy at birth by almost 4 years, and elimination of cancer by more than 3 years. Other leading causes of death have a much smaller impact”.

    United States Life Tables Eliminating Certain Causes of Death, 1999–2001 (2.1 MB)

    1. Anon says:

      Cure cancer and people will only die of more costly diseases like Alzheimer’s. We all have to die once, and cancer is still one of the cheapest ways to go, so total healthcare costs would double even if we could totally cure cancer for free.

      1. loupgarous says:

        The stats show that most Americans 65 or over are co-morbid with Alzheimer’s disease or another dementia and chronic kidney disease, CHF, COPD, CAD, stroke, DM and cancer. Sooooo…. curing cancer and/or heart disease actually saves the nation money, because those conditions at their most disabling also require assistive care of varying degrees, hospitalization, etc.

        We all owe God a death, to be sure. Alzheimer’s may be a preventable illness; even if that’s not the case, treatments are improving for it. I don’t buy into the “immortal American” thesis; Americans have no duty to die on someone else’s schedule to make the social care budget numbers come out right. Shaming people 65 or over for not going gently into that good night is shaming the wrong folks.

        1. Hap says:

          Alzheimer’s in particular is likely to lead to lots of long-term care costs and disruption for everyone involved (both the person who has it and their family). It also makes what life people have with it much less valuable. Getting rid of it would be beneficial to both individuals and society.

          The desire to have people die earlier of less costly diseases also assumes that people don’t have anything to learn or impart in the extra time. That also seems to be the conclusion of management (why employ expensive older people that know stuff we don’t care about and don’t want to pay for?), but it is probably one of their many misjudgements.

          1. Dr CNS says:

            Reminds me of “fail early, fail cheap”….
            I think i have read that expression befoew in this blog.

          2. Anon says:

            An aging population with more people in retirement consuming more healthcare costs is a sick and dying one, whichever way you look at it. It smells like a flask full of dying bacteria.

          3. Hap says:

            If you do it right, though, those old people might be able to do useful things. Everyone gets old. If being old means you know more stuff and can do useful things, not only aren’t you costing society money or utility, but society gains from the deal (probably – population issues?).

            I don’t think Logan’s Run should be taken as a “how-to” guide.

          4. Anon says:

            Most of the retired people I know “contribute” to society by pottering around their garden, watching TV, spending their retirement abroad, staying in a care home, or being treated in hospital. Certainly not working productively, but occasionally looking after grandchildren.

          5. Hap says:

            Even if that’s all you get from the, there’s also the idea that if people didn’t believe they could retire and do things later, they would be more likely to do them now and thus do less work now – it would raise the opportunity costs of work now. Maybe that’s appropriate (you don’t know how long you’re going to live in any case and in what shape you’ll be later), but it would have a cost to society, as well.

  41. Gareth Wilson says:

    It looks like you’re trying to cure cancer.
    Would you like help?

    1. Phil says:

      Go home, Clippy. You’re drunk.

  42. J Severs says:

    “How do you picture science to yourself? ‘Oh, wonder-working steed, build me a palace by morning’ and by morning there is a palace? And what if the problem has been incorrectly stated in the first place? And what if new phenomena turn up?” Alexander Solzhenitsyn, The First Circle.

    1. loupgarous says:

      Good catch. Category errors may be rife in whatever models Microsoft develops of cellular signalling, simply because people who do that research for a living are still discovering how such things work.

      Microsoft could actually be of best service by getting the bugs out of scientific publication. They already have appropriate technology to accomplish that.

      If Microsoft just funded a clearinghouse of qualified peer reviewers for scientific journals, they could spend much less money getting important information where it’ll do the most good for patients, physicians and researchers, while reducing the chances of another “Crap. Courtesy of a Major Scientific Publisher” moment.

  43. J Severs says:

    Also, when all you have is a hammer, everything looks like a nail.

  44. Not A Chemist says:

    While there are some reasonable things in the source material at Microsoft (I am sure they can automate a search of the literature, and machine recognition of patterns in scans or film seems totally doable), there was also this: “one approach is to create a kind of molecular computer that you would put inside a cell to monitor for disease. If the sensor detected a disease, it would actuate a response to fight it”. Which is pretty much saying they can build a better immune system than nature has over the last few billion years.

    And they seem to understand that, since they also say “If [they get] to a point where they understand those systems as well as we understand how to make Microsoft Word run on a PC, they might be able to equip the immune system to mount a powerful response to cancer on its own.” Yes, that is correct.

    My estimate of the probability of success for them understanding cancer in the next ten years has dropped from tiny to exactly zero. I would have to agree that one should discover first and publish second in cases like this.

    But they could still make a boatload of money on some of the more realistic goals.

  45. Mol Biologist says:

    IMO Microsoft is be able to do it if only able to resolve a complexity proposed by Murray Gell-Mann. And to analyze tremendous amount of information. Intellectual efforts must be harnessed to uncover patterns which still hidden.
    There are a lot common features in the process of evolution for human languages and for genomes. Charles Darwin described it for the first time as the existence of parallels between the evolution of species and human languages (The Descent of Man).
    So, kind of agree with Derek it “just grew” in a process, and genomes and languages are formed by pressure of vital needs.

  46. Dr. Manhatten says:

    I just discovered the entire Microsoft announcement was an inadvertent automatic Microsoft Word spelling correction. The announcement was that that Microsoft was going to cure “canker sores”, which was corrected automatically to “cancer, for sure”.

    1. duane schulthess says:

      For context, we may all want to consider the success of Microsoft’s last technology adventure off-piste…

      “Microsoft’s Chatbot ‘Tay’ Just Went on a Racist, Misogynistic, Anti-Semitic Tirade”

  47. jrm says:

    as a computer scientist who has been working in cancer research for a while, I put as much stock into this claim as into Microsoft’s claim in the early 2000s that they would solve spam within 2 years…

  48. Blunderbuss says:

    I thought I couldn’t like this blog more and you’ve done gone and referenced Neuromancer. The perfect analogy to Microsfoft’s plan is found in Primo Levi’s Nitrogen chapter in “The Periodic Table” where a large quantity of chicken manure was attempted to be converted to the dye alloxan, “The shit remained shit…”

  49. Isaac Yonemoto says:

    As someone who has designed (by hand) enzyme mutations that improved enzyme activity ( more than once: (, and also done extensive contract coding in totally unrelated field ( I think the DNA/computer analogy isn’t totally useless (it’s guided some of my – successful – decisions)… But otherwise great article!

    1. Anon says:

      The cell and it’s machinery is so complex and the information that its genome stores may be beyond the limits of our technology to decipher… can be seen here where people have been thinking to use DNA to store the info that our computers can’t store:

      How could you retrieve the info of each of theses cancerous cells/types hold and store it, analyze it and then come up with a programming that can fix it?….

  50. Daniel Barkalow says:

    It’s understandable for Microsoft to have a poor understanding of biology, but they’re showing a poor understanding of computer science as well. Debugging an arbitrary computer program is much harder than curing cancer (in that we’ve proved it is impossible, where as cancer might be possible to cure). You can debug programs that were designed to work for a logical reason, but if you’ve got a program that randomly (or due to evolution) happens to mostly work, there is no technique which has any realistic chance of tweaking it to work completely.

  51. Patrick Sweetman says:

    Come on Derek, don’t be a tease, tell us what you really think.

    PS. I think that this essay could be put to music (Alice Cooper) as a prog-rock anthemn.

  52. nmr.wars says:

    Hi Derek, actually there is a decent analogy between the complexity of cancer to coding. It’s like looking at compiled code. In about 10 years’ time Microsoft will have as good an understanding of cancer biology as my grandma has of an .exe file opened in notepad…

  53. loupgarous says:

    “. . .The programming principles and tools group has already developed software that mimics the healthy behavior of a cell, so that it can be compared to that of a diseased cell, to work out where the problem occurred and how it can be fixed” reminded me of John Conway’s computer game “Life,” written back when men were men, women were women, and apps were called “programs.”

    Granted, Conway was modelling life at the “just-multicellular” level, where the modelling never got inside the cell membrane, but just mimicked what you’d see in a Petri dish. But that sucker would run on 2Kb of RAM on a Sinclair ZX-series microcomputer..

    I’m guessing the Microsoft code that “mimics the healthy behavior of a cell…” does its mimicking in a terabyte or so of RAM and server space.

  54. Anon says:

    I’m getting into carbohydrate cancer immunology.

    No one knows what is going on with that. At all. Maybe everything (they are highly conserved) maybe nothing. It is really impossible to model (and we are modelling…) what is going on at this multivalent, self-organized, rapid turn-over communication level. Biology is way to complex with way too many feedback loops.

    It’s not like a program. It is more like all possible programs are constantly running, you can never see the code, just the output (you can see the binary but you are not sure what coding language is used), you can’t pause it, and all the programs are inter-referential on EVERY SINGLE LINE. With every program influencing every other program constantly and at varying non-linear levels.

    Good luck with that in 10 years. And that is when the system is working properly.

  55. Cancer Sucks says:

    I admit that I did not read every comment, but mostly scanned through them. I get the impression that most of the people that left comments are, in some way, in the medical research field. I can appreciate not wanting your job, or life’s work, trivialized by someone that you feel knows less than you. It is completely understandable.

    I work in I.T. at a pharmaceutical and see trained chemists using computers, instruments, and software every day in their jobs. At every stage in this interaction there was a software engineer that developed based on a user requirement spec from the medical community. This made the job of the chemist easier and, hopefully, less likely to have errors. So, the software engineer didn’t solve the chemists problem – but, gave them tools that they can do this in a more efficient manner. I see that as a good thing.

    I lost someone to cancer last year – I saw treatment after treatment fail and the person I knew slowly morph into a shell of the person I loved. I don’t think this is an easy problem to solve, I don’t know if it is ever going to be completely solved – we might as well be trying to cure ‘death’. What I do know is the more people working on a problem ‘can’ be a good thing. It’s not always good to have to many ‘chefs in the kitchen’, but sometimes a second set of eyes – or another billion dollars or so – can help.

    Bottom line is I don’t care who solves the cancer problem – as long as it is done we all will win. Maybe this announcement from Microsoft will motivate all of the researchers to solve this before the ‘ignorant software engineers’.

    1. loupgarous says:

      I’m a retired programmer from Big Pharma (my job title was “senior consulting analyst”). I wrote and designed programs to turn data from clinical research forms into statistically-tested tables and tabulations useful to my employers to get their new drugs approved for sale, and to monitor the safety of drugs after that point. I’m medically retired, by the way, and have been fighting cancer for eighteen years now.

      So your thesis that I just don’t want my job taken away from me by Microsoft, or that I’m oblivious to the need to solve “the cancer problem” is fallacious on both points.

      I’m in a position to be knowledgeably critical (to a limited degree, anyway) of anyone from the world of computer programming who has huge financial and computing resources but a very spotty record of quality control in their production-level code (Microsoft has never gotten away from making their paying customers their beta-testers) who blithely announces they’re going to apply their expertise to solving problems that people who treat cancer for a living are struggling with.

      The Greek term for that is hubris, which here in the South, we call it “bullshit”.

      Those of us who’ve used Microsoft products since their first DOS would be happy if they just produced an OS that worked well in its first release. We’re not sanguine (a word that works well in both meanings, when you contemplate a company most famous for selling buggy code purporting to tell the oncology community what they’re getting wrong) about Microsoft being one bit better at modelling cellular signaling than they are at making OSes that don’t require quad-core CPUs to work well.

Comments are closed.