5 ms·

"The best minds of my generation are thinking about how to make people click ads. That sucks." [0] The reason why this article resonates with many readers is t

by jal278 12y ago

"The best minds of my generation are thinking about how to make people click ads. That sucks." [0] The reason why this article resonates with many readers is that it expresses frustration with the triviality of how our economy allocates resources to start-ups. The promise of tech is huge, but following the money often leads to optimizing ads, machine learning algorithms competing with each other in HFT, or other applications with little net benefit to society. It's not that EVERYONE should be working on the big problems -- just that the incentive systems seem to lean towards triviality over deep human benefit. [0] http://www.businessweek.com/magazine/content/11_17/b4225060960537.htm http://www.businessweek.com/magazine/content/11_17/b42250609...

rfrey 12y ago
That quote is dripping with hubris. There's a few high profile cases where it's true -- Ng at google and now Baidu, for example -- but I guarantee that the bulk of the "best minds" of the generation are not programmers at Google or Facebook.
- jjoonathan 12y ago
  You have a point, but it's worth mentioning that the fundamental observation behind the complaint is still sound: incentives are rigged to guide smart people towards pursuits of dubious utility compared to what they would otherwise choose to work on. Let's look at a rough picture of incentives for a new graduate: 1. Helping the poor: $0 (maybe $12k if you count a fellowship) 2. Trying to Cure Cancer: $25k/yr, bump to $40k after 6 years 3. Engineering Medical Devices, Airplanes, etc: $60k/yr 4. Trying to Build the Next Twitter: $100k/yr-$150k/yr 5. Helping Rich People Game the System to Get Richer: $150k/yr, $300k/yr after a few years if successful I have placed these in decreasing order of social value per my own judgement. Note that salary has precisely the opposite trend. The probability of that particular relative order occurring by chance is 1 in 120. I could add other jobs to the list to drop the probability even further. In some cases the market is conveying valuable information. If we have enough !Software Engineers, it makes sense to depress their wages as a way of discouraging people from entering those fields. Similarly, it makes sense to encourage smart people to pursue roles with a lot of leverage (in the value creation sense, not the bargaining sense), which is a plausible excuse for sending smart people into finance (not one that I buy, but it's plausible). But in other cases we can identify the information that the market is conveying and pass judgement with the opposite conclusion. By design, the market rewards those who feed and clothe wealthy people and punishes those who feed and clothe the poor and penniless. In this instance the market's reasoning is transparent enough to judge and few would rule in its favor ("necessary evil" is the best I've heard, and then only in the context of the circular "it's the best system we've tried" argument). Another negative example would be medical research. Next time someone you care about is sick or dying and has run out of options, reconsider the wisdom of our system in spending ~3% of all medical expenditures on research. Also remember that much of that is governmental -- if the market had its way, that figure would be even lower. The market's reasoning is transparent: a doctor, insurance company, or drug company can threaten to withhold the service or good they provide in order to capture a fraction of the value they create. A researcher can't do that, at least not effectively (their product only appreciates in value once they no longer control it). So they are systematically undervalued and the options that your loved one has are artificially reduced as a result of this. Market-based incentives have huge problems. The million dollar question (many trillion dollar question) is how to fix them.
timr 12y ago
"It's not that EVERYONE should be working on the big problems -- just that the incentive systems seem to lean towards triviality over deep human benefit." Yes, exactly. And even more importantly: we can fix it, but we choose not to. I know brilliant people with extremely specialized training in chemistry, physics, biochemistry, and so on. People who should be developing new drugs, or high-speed trains, or improving solar energy, or creating new ways to generate hydrogen or ethanol from plants. You know what? They can't get jobs in their field of training, and if/when they do, the jobs are crap. The sad truth is, it's not (currently) economically rational to devote your youth to these high-minded pursuits. In all likelihood, you will be rewarded with a lifetime of insecurity and low income, followed by an uncertain potential of a comfortable life in old age. This is a societal problem. We're currently spending all of our resources on short-term wealth accumulation and neglecting longer-term technological development. We could stop it with the right set of financial incentives, but we don't.
- JoshuaDavid 12y ago
  Who is "we"?
  - TeMPOraL 12y ago
    The humanity.
onewaystreet 12y ago
Ad money is what's paying for all the interesting projects and acquisitions going on at Google right now. If not directly a net benefit to society, it has been one indirectly.
- shas3 12y ago
  Paraphrasing Peter Thiel [1], this is a stochastic approach ("unplanned statistical view of the future") to technological progress. Throw darts at ad clicks and hope that it cures cancer. The opposite approach is the 'Manhattan Project'-way, where you throw the best minds in a field to work on a particular project. To some extent, this is going on in areas like mathematics (Langlands Project), Human Genome Project, NIH BRAIN Initiative, etc. In the long term, it is unclear if the 'stochastic approach' is deterministically effective in solving the 'Big' problems. [1] http://www.the-american-interest.com/articles/2012/02/01/a-conversation-with-peter-thiel/ http://www.the-american-interest.com/articles/2012/02/01/a-c...
digz 12y ago
Why is helping people find products and services that will enrich their lives an inherently bad thing?
- gress 12y ago
  Advertising is a service designed to help corporations exploit psychology to induce people to buy their products. What you are describing is more like a concierge. Hint: The person who pays is the person who's interests are being served.
  - digz 12y ago
    Reaching consumers is hard, even if you have the superior product. Facilitating such a transaction should make one proud. The 'psychological exploits' are a marginal component, but again, why is it inherently bad for someone to feel good about using product x over product y, even if they're exactly the same?
    - forgottenpass 12y ago
      Reaching consumers is hard, even if you have the superior product. Facilitating such a transaction should make one proud. It's entirely an accident if an advertizing service facilitates that transaction. They're not consumer reports, they're boosters for whoever pays.
    - gress 12y ago
      Are you suggesting that advertising never influences people to make choices that are not in their best interests?
- Houshalter 12y ago
  Advertisements are more manipulative than informative. But the bigger point was that they are far less of a benefit to humanity than other things, yet they get far more resources allocated to them.
- RogerL 12y ago

by jal278 12y ago

rfrey 12y ago

That quote is dripping with hubris. There's a few high profile cases where it's true -- Ng at google and now Baidu, for example -- but I guarantee that the bulk of the "best minds" of the generation are not programmers at Google or Facebook.

jjoonathan 12y ago
You have a point, but it's worth mentioning that the fundamental observation behind the complaint is still sound: incentives are rigged to guide smart people towards pursuits of dubious utility compared to what they would otherwise choose to work on. Let's look at a rough picture of incentives for a new graduate: 1. Helping the poor: $0 (maybe $12k if you count a fellowship) 2. Trying to Cure Cancer: $25k/yr, bump to $40k after 6 years 3. Engineering Medical Devices, Airplanes, etc: $60k/yr 4. Trying to Build the Next Twitter: $100k/yr-$150k/yr 5. Helping Rich People Game the System to Get Richer: $150k/yr, $300k/yr after a few years if successful I have placed these in decreasing order of social value per my own judgement. Note that salary has precisely the opposite trend. The probability of that particular relative order occurring by chance is 1 in 120. I could add other jobs to the list to drop the probability even further. In some cases the market is conveying valuable information. If we have enough !Software Engineers, it makes sense to depress their wages as a way of discouraging people from entering those fields. Similarly, it makes sense to encourage smart people to pursue roles with a lot of leverage (in the value creation sense, not the bargaining sense), which is a plausible excuse for sending smart people into finance (not one that I buy, but it's plausible). But in other cases we can identify the information that the market is conveying and pass judgement with the opposite conclusion. By design, the market rewards those who feed and clothe wealthy people and punishes those who feed and clothe the poor and penniless. In this instance the market's reasoning is transparent enough to judge and few would rule in its favor ("necessary evil" is the best I've heard, and then only in the context of the circular "it's the best system we've tried" argument). Another negative example would be medical research. Next time someone you care about is sick or dying and has run out of options, reconsider the wisdom of our system in spending ~3% of all medical expenditures on research. Also remember that much of that is governmental -- if the market had its way, that figure would be even lower. The market's reasoning is transparent: a doctor, insurance company, or drug company can threaten to withhold the service or good they provide in order to capture a fraction of the value they create. A researcher can't do that, at least not effectively (their product only appreciates in value once they no longer control it). So they are systematically undervalued and the options that your loved one has are artificially reduced as a result of this. Market-based incentives have huge problems. The million dollar question (many trillion dollar question) is how to fix them.

timr 12y ago

"It's not that EVERYONE should be working on the big problems -- just that the incentive systems seem to lean towards triviality over deep human benefit." Yes, exactly. And even more importantly: we can fix it, but we choose not to. I know brilliant people with extremely specialized training in chemistry, physics, biochemistry, and so on. People who should be developing new drugs, or high-speed trains, or improving solar energy, or creating new ways to generate hydrogen or ethanol from plants. You know what? They can't get jobs in their field of training, and if/when they do, the jobs are crap. The sad truth is, it's not (currently) economically rational to devote your youth to these high-minded pursuits. In all likelihood, you will be rewarded with a lifetime of insecurity and low income, followed by an uncertain potential of a comfortable life in old age. This is a societal problem. We're currently spending all of our resources on short-term wealth accumulation and neglecting longer-term technological development. We could stop it with the right set of financial incentives, but we don't.

JoshuaDavid 12y ago
Who is "we"?
- TeMPOraL 12y ago
  The humanity.

onewaystreet 12y ago

Ad money is what's paying for all the interesting projects and acquisitions going on at Google right now. If not directly a net benefit to society, it has been one indirectly.

shas3 12y ago
Paraphrasing Peter Thiel [1], this is a stochastic approach ("unplanned statistical view of the future") to technological progress. Throw darts at ad clicks and hope that it cures cancer. The opposite approach is the 'Manhattan Project'-way, where you throw the best minds in a field to work on a particular project. To some extent, this is going on in areas like mathematics (Langlands Project), Human Genome Project, NIH BRAIN Initiative, etc. In the long term, it is unclear if the 'stochastic approach' is deterministically effective in solving the 'Big' problems. [1] http://www.the-american-interest.com/articles/2012/02/01/a-conversation-with-peter-thiel/ http://www.the-american-interest.com/articles/2012/02/01/a-c...

digz 12y ago

Why is helping people find products and services that will enrich their lives an inherently bad thing?

gress 12y ago
Advertising is a service designed to help corporations exploit psychology to induce people to buy their products. What you are describing is more like a concierge. Hint: The person who pays is the person who's interests are being served.
- digz 12y ago
  Reaching consumers is hard, even if you have the superior product. Facilitating such a transaction should make one proud. The 'psychological exploits' are a marginal component, but again, why is it inherently bad for someone to feel good about using product x over product y, even if they're exactly the same?
  - forgottenpass 12y ago
    Reaching consumers is hard, even if you have the superior product. Facilitating such a transaction should make one proud. It's entirely an accident if an advertizing service facilitates that transaction. They're not consumer reports, they're boosters for whoever pays.
  - gress 12y ago
    Are you suggesting that advertising never influences people to make choices that are not in their best interests?
Houshalter 12y ago
Advertisements are more manipulative than informative. But the bigger point was that they are far less of a benefit to humanity than other things, yet they get far more resources allocated to them.
RogerL 12y ago