Author Archives: Mike White

Reupping:Why reproducibility initiatives are misguided

I’m reposting this two-year old piece, because it’s worth reminding ourselves why exact replication has, with minor exceptions, never been an important part of science:

In my latest Pacific Standard column, I take a look at the recent hand-wringing over the reproducibility of published science. A lot of people are worried that poorly done, non-reproducible science is ending up in the peer-reviewed literature.

Many of these worries are misguided. Yes, as researchers, editors, and reviewers we should do a better job of filtering out bad statistical practices and poor experimental designs; we should also make sure that data, methods, and code are thoroughly described and freely shared. To the extent that sloppy science is causing a pervasive reproducibility problem, then we absolutely need to fix it.

But I’m worried that the recent reproducibility initiatives are going beyond merely sloppy science, and instead are imposing a standard on research that is not particularly useful and completely ahistorical. When you see a hot new result published in Nature, should you expect other experts in the field to be able reproduce it exactly? Continue reading

The FDA Should Only Approve Drugs Backed By Science

Here’s some Inauguration Day reading for you:

For half a century, the FDA has regulated drugs on the premise that science should show that they’re safe and effective before drug companies get to sell them. Before modern drug laws, companies filled the market with ineffective products, backed by no evidence that they worked.

As I write this week in Pacific Standard, a couple of candidates for Trump’s FDA seem to think that we should go back to that era before modern drug laws. Jim O’Neill, a venture capitalist who invests in biotech argues that as long as companies can show a drug is safe, the FDA should let patients take it “at their own risk,” regardless of whether that drug is useless.

And biomedical engineer/biotech executive Balaji Srinivasan thinks that, rather than testing drugs with clinical trials, people should just rate them the way they rate their Uber drivers. Given that people believe all sorts of insane things about what makes them healthier, this is not likely to be a way to rigorously learn whether a drug actually does something for the patients who buy – and whether its not just the drug company ripping people off.

In my piece, I explain exactly why these ideas would be bad for you. But the overarching theme is this: the problem with the two candidates, and their associate Peter Thiel (who is advising Trump on the FDA) is that they see drugs and biotech from the view of investors and startup executives. These are people who hear promising, brilliant medical ideas all the time from scientists and entrepreneurs, people who want to take fledgling ideas and turn them into therapies. That’s great, but in the end, most of these promising, brilliant ideas will in fact be wrong – and that’s why we need the FDA to protect us by weeding out the failures.

We Sell Our Attention Too Cheaply

The message of Tim Wu’s The Attention Merchants is that we should more carefully consider what we get in return when we let companies sell our eyeballs to advertisers. In fact most we don’t consider it at all. If I’ve failed to turn off Twitter or my mail app while working, inevitably, barely thinking about it, I click away from the task at hand to quickly check my inbox or Twitter feed. I’m lured away from productive work by something that is literally optimized to capture my attention.

In the case of Twitter, as all of us these days are aware, the salable product isn’t the app, it’s your attention. Twitter, Facebook, Google, LinkedIn, Instagram, are nominally free because these companies are  what Tim Wu calls attention merchants, capturing your mental space and selling it to advertisers. In exchange, we get pretty good free email, a way to stay in touch with old high school friends, and a chance to argue politics with idiotic strangers who we’ll never meet in real life.

The attention merchant model isn’t new – Wu traces it back to the first tabloids in New York, which sold their papers below cost, hoping to make money through advertisers. To make this business model work, tabloids like the New York Sun were the Buzzfeeds of their day, trafficking in the 19th century equivalent of clickbait.

In the case of Gmail or Facebook, yeah, we know they’re selling our data, but we think the trade is worth it. The same is true with broadcast TV: the Superbowl is free to watch, and in exchange, we sit through some high-budget, clever ads. What’s not to like about this deal?

Wu argues however that we’re selling ourselves too cheaply. Without blinking an eye, we let advertisers vie for our attention in our most personal and protected spaces: our homes, our bedrooms, our schools. At one time, as Wu documents, the idea that one would let a stranger pitch a sale to the family sitting around the dinner table was outrageous. No longer, as we eat in front of our TV’s and phones.

Not only have the attention merchants completely invaded every last corner of our personal space, but their business imperatives have also distorted the content we get. We think we get a great app or great TV shows in exchange for looking at a few ads. Wu shows however, we put up with content that is not driven by the creative talents of directors, screenwriters, performers, journalists, or editors, but rather by the talents of those whose business is to commoditize human attention – and the content sucks as a result. As examples, Wu cites the rigged quiz show craze of the late 1950’s, or the crappy reality TV of the aughts. (Interestingly, Wu argues that we’re getting great television now, thanks to companies like Netflix that, rather than selling attention, sell subscriptions.) One could also cite the deluge of completely fabricated news stories during this past election.

One question that Wu didn’t seem to me to have a good answer to is, what’s the alternative? Take journalism for example, perhaps the most important locus of tension between the attention merchant business model and the critical societal need for content that does not actively mislead or exploit readers or viewers. Few newspapers these days can survive on subscriptions – so how do they stay afloat without selling out their readers? Cable news has clearly gone all-in on the attention merchant model, pursuing sensational but trivial stories to keep viewers glued to the screen.

I got bogged down in a few places in this book – Wu’s critiques of specific social media apps is very detailed – but, overall, Wu gives the reader a deeply engaging history of attention merchants and advertising. He ties together the first tabloids, the original snake oil salesman, and the tortured history of AOL.

I don’t think I’m as pessimistic as he is, but we absolutely should take seriously his call for a “human reclamation project”, creating protected zones in our lives “for our own consciousness and mental space.” While we unquestionably benefit from many of the “free” things we get from attention merchants, we’ve failed to adequately reckon with the consequences to human civilization of the commodification of our mental space. “What are the costs to a society of an entire popluation conditioned to spend so much of their waking lives not in concentration and focus but rather in fragmentary awareness and subject to constant interruption?” Wu asks.

Seeing how we’ve just come through an election dominated to an unprecedented degree by fake news, propaganda, and other ridiculous distractions from the very high stakes that were on the ballot, we’re about to learn the answer.

You can hear Tim Wu talk about this and more on the Ezra Klein show, where I first heard about this book. And yes there are ads on this blog, but I guarantee you they’re well worth the content you get here.

Genomics and the Power of Public Health

On a bad day in the lab, we sometimes joke that if we really wanted to help save lives, we’d forget about molecular biology and go help people quit smoking. Relatively simple public health efforts – clean water, washing your hands before moving from one sick patient to another, basic vaccines – generally save many more lives that the cures that come out of the high tech stuff we do in the lab. Cancer immunotherapy may turn out to be a major advance in cancer treatment, but we’d reduce cancer even more if we could get everyone to quit smoking, lose weight, and stay physically active.

Genomics, whose near-term medical benefits have been the subject of a lot of hype, may turn out to be a high-tech, scientifically complex effort that actually does to have a big impact on people’s lives. As I discuss in my Pacific Standard column this week, part of that will be the long-term medical benefits that grow out of a better understanding of biology. But a more dramatic – and more near-term – impact may be how genomics changes public health. As some sort of genome analysis becomes a routine part of normal medical care, genetics will be integrated with other public health screenings (like testing your cholesterol), which, as two recent studies show, could have a big impact on avoiding preventable consequences of common diseases. Once exome sequencing is cheap enough, there could be a possible benefit of combining genomic screenings using existing medical knowledge – we don’t need to wait for distant future discoveries.

The takeaway is that policy and the infrastructure of the healthcare system, and not science, may soon be the rate-limiting step for realizing the medical benefits of genomics in some cases. Physicians, insurers, hospitals, and the health care system in general is utterly unprepared to handle the kinds of genomic data that could, in the near future, improve routine medical care.

A Long View of the Scientific Revolution

wotton Been eager to get this doorstop of a book, which sat on my shelf for most of 2016:  David Wotton’s The Invention of Science. I’ll have more to say later, but for now, a few pages in, it’s clear that Wotton really takes the long view on science. “It is far too soon to say” how the Scientific Revolution will turn out:

But since 1572 the world has been caught up in a vast Scientific Revolution that has transformed the nature of knowledge and the capacities of humankind. Without it there would have been no Industrial Revolution and none of the modern technologies on which we depends; human life would be drastically poorer and shorter and most of us would live lives of unremitting toil. How long it will last, and what its consequences will be, it is far too soon to say; it may end with nuclear war or ecological catastrophe, or (though this seems much less likely) with happiness, peace and prosperity.