Hacker News Reader

Comment by projectramo | Hacker News Reader

And you know what: we can't dismiss him.

The interesting thing here is that we can't just say that's not true.

I am not making a larger point about solipsism or our inability to know reality. I mean the much narrower point of not being able to tell -- from the press -- whether a company actually has a technology that works.

Think about the implications that has for investing in public stocks.

Let's ignore the cases of fraud like Enron etc., and focus on an ordinary company that honestly reports what it is doing.

EDIT: I have followed the Theranos story, so I know about the legal actions etc.

What I can't tell is if there is poor science at the heart of it, or it is a case of poor scaling and application of good science.

View on HN

> I mean the much narrower point of not being able to tell -- from the press -- whether a company actually has a technology that works.

There has always been a simple solution to this: it is the company's own job to clearly demonstrate that it has a working technology.

If you view it that way, then the press's inability to prove it one way or the other doesn't matter because the responsibility lies with the company. In fact, Theranos was quite unusual among lab companies from the beginning in that it didn't follow the path other companies followed in showing they had a working technology. Anything less than a clear proof by the company, and the subsequent confusion is the company's own fault.

To demonstrate it to whom?

> I mean the much narrower point of not being able to tell -- from the press -- whether a company actually has a technology that works.

There have been empirical actions which resulted based on Theranos's errors, including FDA/SEC investigations and Walgreens revoking their partnership. Those aren't media spins.

But didn't Walgreens have to test the technology to agree to the partnership in the first place?

There is a difference between the tech that doesn't work and tech that cannot work.

The former will work eventually, but the latter can never work.

Can you tell which it is?

> But didn't Walgreens have to test the technology to agree to the partnership in the first place?

Famously Walgreen's didn't do due diligence.

> There is a difference between the tech that doesn't work and tech that cannot work.

It was already pretty well known that blood concentrations vary enough in the body that you need large samples to get reproducible results. Life science VCs all passed on this deal. The fundamental science was not there.

Note: I have worked in pharma and designed successful clinical trials.

>>I mean the much narrower point of not being able to tell -- from the press -- whether a company actually has a technology that works.

http://www.wsj.com/article_email/theranos-voids-two-years-of... ??

I don't know what point you are trying to make, but Theranos is not some reality-bending exceptional case. You might say it's possible he is correct, but, yes, one can dismiss "There's nothing wrong with Theranos." There have been documented wrongs that nobody has tried to refute.

In your opinion, has it been documented that the technology doesn't work or that they couldn't apply the technology properly in the field?

In other words:

1. Is it fraud and poor science? 2. Is it good science but still immature?

That is the key distinction.

Your question attempts to make a distinction without a difference. The Theranos machine does not work because nobody knows how to make whatever theory they had work in practice. Moreover, Theranos's management knew that as they were selling unreliable lab results.

Well, after the FDA shut down many of the company's operations, and Walgreens pulled out of their contract to use Theranos services, one might suspect that their technology does not actually work.

If it did work, it would be straightforward for them to demonstrate that it did.

Didn't it have to demonstrate to both Walgreens and the FDA that it was working before?

I can't tell if it never worked, or if something prevented it from scaling.

Reply on news.ycombinator.com