Hacker News Reader

Comment by maxharris | Hacker News Reader

the hard part of it

This is absolutely false. I agree that discovery isn't easy, but you know what's even harder? The other 90% of the work needed to finish it and make it a reliable product. Chomsky trivializes the work of synthetic organic chemists, the engineers that set up and keep production lines going, researchers that test the drugs, and so on.

In software terms, this is like saying that all the hard work is done because someone banged out a prototype over the weekend. It's full of bugs and doesn't have the features people want yet.

If you believe that Chomsky's right, try running a startup while holding his idea. You'll fail, or you'll discover that it's not easy and reject his idea.

View on HN

He's talking about research. You're talking about development.

None of these things are exactly easy, but at least with development you're building toward something.

Where does one stop and the other begin?

Much of what we call research vs. what we call development is simply an artifact that stems from the way science is funded today. However, science can't (and won't) always be funded by giant government bodies. (To say that it will assumes that the way things are can go on forever, but history shows that civilizations have lifespans just like everything else.)

Pharmaceutical R&D companies have become just development companies because they stop looking for new drugs and are simply repackaging and reformulating existing drugs (e.g. Claritin/Clarinex).

They're doing this because it costs too much to do research and development gives them quicker return on investment.

They'll do their own research when the government stops competing with them in that arena. Until then, why duplicate it when they (and everyone else) has been forced to pay for it anyway? (I'm all for research, but I'm against forcing people to pay for it. The best scientific research is intricately tied into industry anyway. For example, where would biochemistry be without cyclotrons, radiolabeling, glassware and chemical reagents? All of that was possible because of private industrial work. Ever heard of Peter Mitchell?)

Second, pharmas aren't free to develop whatever they think would be profitable. New drugs cost around a billion dollars to develop because of the FDA, which is ridiculous. There are many rare illnesses for which affected people don't want the "safety" the FDA imposes. As long as billion dollar development costs are imposed on every drug, there is no chance that any progress can be made for these people. So it's not the pharmas, it's the regulation.

The other side of that argument is of course that most patients don't want "unsafe" drugs, so that regulation has a purpose.

Is there any way to get a best of both worlds in this scenario? Quicker to-market times for drugs for rare and dangerous diseases, and the safe route for the more common drugs?

Is there any way to get a best of both worlds in this scenario? Quicker to-market times for drugs for rare and dangerous diseases, and the safe route for the more common drugs?

Of course there is. Regulations can be changed, freeing drug companies to research and develop these drugs. In fact, I'm sure that this change - this partial deregulation compromise - is going to happen, someday.

But on what principle should we stop there? Where do you draw the line, and why? If the regulations say "freedom only for drug developers that treat diseases that affect 15,000 people with five years," on what principle do you cut freedom short for the 15,001th patient, or the one who is expected to live five years and a day?

You draw the line where it works best, because it works best.

The goal isn't some nebulous definition of "freedom", where "freedom" is defined as "adherence to conservative orthodoxy regardless of the particulars".

The goal is something that works well.

I disagree.

Who decides "what works best", and how?

Edit: I say that each of us has to do that for ourselves.

Also, I'm not a conservative. I'm anti-religion, pro-selfishness, pro-gay, anti-environmentalism, pro-abortion, pro-capitalism. I'm a radical for reason.

In this conversation, "works better" was defined 2 short comments ago as "easing the regulatory burden for exceptional cases [and in general if possible] while still maintaining safety standards for mundane drugs". Paraphrased.

But you jumped way past "working better" into some unrelated and 100% ideologically-driven conception of "freedom". What are we talking about here, my freedom to get poisoned by cough syrup because I didn't perform my own scientific studies on the matter with a large sample population?

How come freedom only ever applies to the already-rich cutting corners for an extra buck?

How come freedom only ever applies to the already-rich cutting corners for an extra buck?

It doesn't. Yaron Brook just addressed this in a debate held in New York last week:

"The biggest victims of government intervention in the economy, the biggest victims of socialism (as someone who came from a socialist country, where government intervened a lot more than here) are the ambitious poor - the people who could rise up under freedom.

The people who want to work, who want to make something of their lives, who want to succeed, who want to prosper. They are the real victims. They are the ones we should shed tears for when we regulate, when we control, when we put them through public education, when we put them through the whole social mechanism, when we use welfare to institutionalize them into poverty and take away their sense of personal responsibility and personal morality that is so crucial for their development as successful, happy, prosperous human beings. If anything Objectivism should rally around, it's those people. You know, I wasn't rich, I was an ambitious poor guy. I came to this country with nothing. Those are the people who are the real victims of state intervention, and the real benefactors of freedom and capitalism."

This quotes is from the last five minutes of audio: http://www.archive.org/details/FirstPrinciples-Government_Wh...

Well, pardon my phrasing here, but all you're doing here is doubling down on the hand-wavey bullshit in longer form. I could give you a similar academic paragraph from a trotskyist, it would be just as substantive and just as wrong.

To his point, in the last 30 years, we've slashed taxes for the rich and really slashed welfare. Shouldn't the poor be picking themselves up by their bootstraps by now? Or is something a little more robust than "I don't even need to know the details, the problem is SOCIALISM and the solution is FREEDOM" required?

And meanwhile, none of this has anything to do with the very real problem of FDA regulatory reform. You haven't laid out the first reason why I wouldn't be poisoned by a bottle of cough syrup from duane reade or why it should be on me vs the drug company to prove that their cough syrup is safe.

I suppose we could always have a company that was in charge of rating drugs, and the drug companies could pay them for ratings, like AAA and BBB. I mean, it works for the financial industry, right?

You haven't laid out the first reason why I wouldn't be poisoned by a bottle of cough syrup from duane reade or why it should be on me vs the drug company to prove that their cough syrup is safe.

No, just see the other comment right below:

You should be free to do this yourself or to seek the advice of those you've chosen to trust on scientific (or any other) matters.

As to your comment about the financial industry: it's inapplicable because it's not a free market. In a free market, there would be no Federal Reserve, no Fannie or Freddie to distort the market. Immoral firms such as Goldman Sachs would have nothing to take advantage of.

.. and my trotskyist would argue that the Soviets failed because they never achieved "true" communism. He's just as right as you.

Gotta go, but on that whole federal reserve thing, go take a look at the boom/bust cycle in the US prior to and after establishing the federal reserve.

Again, it's not about what you find ideologically satisfying, it's about what actually works in the real world.

EDIT: btw, the fed and fannie/freddy had almost zero to do with the housing crisis. The problem was the multiplied leverage much more than the housing debt itself, and fannie/freddy's rules about loans to minorities only affected a miniscule % of total defaults.

Again, it's not about what you find ideologically satisfying, it's about what actually works in the real world.

Actually, it's got to be both. Being smart is being able to correctly predict things, and that requires both theory and empirical observations. While everyone has some sort of philosophy, some philosophies are better in the real world than others.

Being smart means also knowing that you inherently cannot predict things. You may be better than others at getting stuff right, yet you are inevitably going to be wrong.

Its pretty established that the financial firms doubled down on leverage and then blew up. A free market would have made things worse, faster and earlier. Nothing could save you from people who are lying to you, and have begun to believe their own horse. No amount of informational accuracy will help.

isn't this a contradiction in terms? in a bubble, 'earlier' equates to a smaller burst

because I didn't perform my own scientific studies on the matter

You should be free to do this yourself or to seek the advice of those you've chosen to trust on scientific (or any other) matters.

We have. We've chosen the government, because they're probably more neutral and objective than a private entity, and that's more important than them probably being less efficient.

> In software terms, this is like saying that all the hard work is done because someone banged out a prototype over the weekend. It's full of bugs and doesn't have the features people want yet.

That's the early Google, Facebook etc. for you. Of course most of such ventures fail, but sometimes they get it done and gain traction.

I guess his argument is based on the premise it takes organization on scale of a state (country) to be able to organize broad-sweeping long-term research in areas that don't look promising at first. Without the sight of immediately sellable product or service. Businesses seem to be more oriented towards research that improves incrementally something that's already somewhat known.

I was going to write Bell Labs were one of notable exceptions, but then I realized they got much state funding for research related to military.

Yes, and Chomsky also offered his view on Bell Labs. An off-the-cuff interview: (http://tech.mit.edu/V128/N51/chomsky.html). A bit more detail, if you search for the word "transistor": (http://www.chomsky.info/talks/199805--.htm).

If you're talking about keeping production lines going - this isn't really part of what no-IP would do. Someone would still need to operate and maintain production and distribution - however the competition to do this would drive costs down dramatically.

Also, high margins and no competitive pressure make research and dev methods stagnant - no IP + public research funding and FDA subsidies seems like a more effective system to me.

Chomsky trivializes the work of synthetic organic chemists, the engineers that set up and keep production lines going, researchers that test the drugs, and so on.

Because private companies would be willing to throw money into the wind to fund research. The concept's been proven, the risky part is over; although not necessarily the difficult part.

Blue-sky research is much higher risk. Without the shorter term payoff that corporations seek. It's reasonable to call an endeavor "harder" when it's less certain of success. Particularly when we're talking about real innovations like computers and the internet, versus a company hiring a couple guys to toss together a "minimal viable product" with PHP.

(If that description of corporate productization sounds unfair, I think it's closer to the mark than saying basic research is analogous to "someone banged out a prototype over the weekend".)

When it's able to be productized, that's when corporations come along. This is well known. (http://www.nytimes.com/2005/04/02/technology/02darpa.html)

Blue-sky research is much higher risk.

That wouldn't stop people from doing it privately. For example, biochemist Peter Mitchell (https://secure.wikimedia.org/wikipedia/en/wiki/Peter_D._Mitc...) won a Nobel in Chemistry in 1978 for work he did entirely without government funding, out of a house he owned. (It's not some little thing he discovered, either - if you take an undergraduate course in biochemistry, you'll run across his chemiosmotic hypothesis directly, and it's taught in less detail in a lot of other biology courses.)

You're right, he was an exceptional individual.

Conversely, the impossibility to exclude (copyright and patents) didn't stop the industry from funding the Linux kernel.

But both cases are probably marginal. In a world where one has to earn money to survive, innovation is slowed down, either because of copyright and patents, or the lack of funding.

The trick is to get a world where one doesn't need to earn money.

"The trick is to get a world where one doesn't need to earn money."

At the very least we need to move towards a place where money is not the final arbiter of 'Good'.

"Intellectual property rights has very little to do with individual initiative. I mean, Einstein didn't have any intellectual property rights on relativity theory. Science and innovation is carried out by people that are interested in it. That's the way science works. There's an effort in very recent years to commercialize it, like they commercialize everything else. "

Interesting example. Pete was actually my step-grandfather, and I visited Glynn quite often as a kid. One thing worth bearing in mind is that he was quite wealthy, having inherited a lot of money (from the wimpey family I believe). So while it is clearly true that some blue sky research can occur without government funding, his path may not be all that easily replicated by others

Try running a startup without Linux and the entire open source ecosystem.

I'm a BSD fan myself, but startups are possible without open source. Stack Overflow runs on Windows, for example.

The one thing that startups need is freedom to make what people want so that those people will freely choose to pay the startup (how many startups existed in the USSR in 1970?). OSS is just a nice-to-have once you look at it that way. If it was a necessity, you'd have to explain how there were startups before anything was open source.

Edit: OSS would still exist in some form even if there were no government funding for it. You know how some people say they do it just for fun? I think that's true.

SO is the exception that proves the rule.

Also, another thing. Linux/BSD push the bar way high up and force Windows to try and catch up. So even if you use windows, you're still affected by Linux. In the same way that you're affected (positively) by Apple even if you use Ubuntu.

The operating system is just one component. People had lives and did things before there were operating systems of any kind.

Edit: For example, could you write CAD software without an OS as we know it today? It would be undoubtedly harder, but you could. People did this in the 60s.

I'm not sure how this fits into the discussion, but what you're saying can be used as an argument against IP.

Since people lived just fine before technology, we don't need technology all that much, therefore we don't need IP.

It might be easier to write CAD software without an OS. At least, writing directly to a framebuffer is simpler than using X11 or OpenGL or Win32. Unless I'm missing something, a modern CAD system has to plot each entity twice; once to the display system and once to an internal grid. The internal grid is used for entity selection, possibly culling and other things.

The older AutoCADs (on DOS) apparently plotted the internal grid (hi-res), then resampled it to generate the screen display.

Thus, minor panning and zooming could be accomodated without "regenerating" -- replotting the hi-res grid.

Anyway, a CAD system doesn't really use the facilities of an OS; something like a web server, however, seems to benefit from an OS.

Minor rhetorical note - you used the phrase "exception that proves the rule" incorrectly. You used it to mean "the single exception to a general rule", with "open source software is necessary for a successful startup" being the rule, and Stack Overflow being the exception.

The phrase actually means that the existence of a specific rule implies that the negation is true for circumstances not covered by the rule. So, if half of the blocks in a city have "parking prohibited" signs, and no such signs elsewhere, that "proves" that parking is in fact allowed on the other blocks, since the signs are excepting the general rule. In other words, why would there be a sign if it was understood that there is no parking?

Sorry to make such a big point out of a small matter, but we already lost one good phrase ("Begging the question") to widespread improper usage. Wikipedia has more info, including the origin of the phrase: http://en.wikipedia.org/wiki/Exception_that_proves_the_rule

The rule is like: It's really really hard to make startups without the open-source Unix-based systems/tools

The point is that the phrase "exception that proves the rule" should only be used when talking about man-made rules and laws, not inductive hypotheses. You don't get to use evidence against your hypothesis as evidence for your hypothesis!

Put another way, when used correctly, "the exception that proves the rule" is a valid rhetorical phrase regardless of how many exceptions there are. Whereas if there were 500 successful startups using non-open-source tools, you wouldn't say that they were the exceptions that proved the rule.

> You don't get to use evidence against your hypothesis as evidence for your hypothesis!

But you do. That's the point. It's not an absolute law. I'm sure you can find a great application written in Visual Basic. This doesn't make Visual Basic a great choice over Python.

If you could list 50+ successful startups running on Windows, then you could say there's a strong evidence.

And even then, this only proves that Windows is as good as Linux. It doesn't counter the original point about Linux being great despite not having a business model centered around IP.

I completely agree with your overall point - that open source Unix-type tools are better for startups, and that Windows-based tools aren't a good fit. I was simply saying that you used a rhetorical phrase incorrectly.

Are the significant Linux/OSS projects publicly funded?

That depends on exactly what you mean by that. The funding sources are quite diverse. Most often, they start out with no funding at all and then, once they solve problems of interest to enough people, they start to attract both government and private sponsors.

But yes, there are government grants. Here's the first example I could find, one for developing a secure Linux desktop:

http://www.tuxmachines.org/node/1153

They don't depend on the IP model; that's what matters

Reply on news.ycombinator.com