[adamt]

I will freely admit I know very little about the pharmaceutical world, but my (perhaps dumb) question is what scope is there to disrupt these typical costs.

For example it used to be that space programs cost billions, yet India has just put a Mars orbital in place for $74m. I am old enough to remember when it was the received wisdom that people believed an e-commerce website would cost $1m, yet many startups launch their first site with less then 0.1% of that.

How much of that $1bn-$2bn typical cost is open to similar levels of disruption?

[Fomite]

You can make way cheaper drugs than $1-2 billion (fidaxomicin was developed by Optimer for 200some million) by being very, very picky about your targets, and not failing.

But it's not a trivial thing to disrupt, and there are some costs that are just out and out fixed (clinical trials are spendy).

[Kalium]

If you can always be right about every drug, every time, at every stage, then perhaps quite a bit. Of course, multiple years of trials is still going to be very expensive.

A lot of the costs seem to be linked to the fact that failures are common at every stage. Which is why there is a lot of time, money, and energy being put into reducing failure rates.

There's not an easy "insert disruption here" point.