This article sounds a bit incoherent and under-researched.
Software “nerds” were the unsung heroes from the very early days of spaceflight. A good book on the involvement of software engineers during the Apollo Program is Mindell’s “Digital Apollo”.
I agree on the first part, hard disagree on the second part.
Publicly funded research part is on the characterization of the disease and the mechanisms, not the actual design of the drugs themselves. A lot more intellectual effort goes into the actual development of the compound that can safely treat the patients.
As I’ve said elsewhere, the majority of the cost of developing a drug is from clinical trials, to make sure that the drug is safe and effective for human consumption. Unless we commoditize the medical services in US, this cost isn’t coming down. In fact, it’s going up.
Asking for lower price is simply asking for fewer whizzy new drugs. If our medical system price goes down, so will drug price.
More people need to be aware of how the drug development costs are distributed. It’s the clinical trials, not R&D, which contributes to astronomical cost, and that price is emphatically NOT set by the pharmaceutical companies that the public loves to point their fingers at.
It’s a simple math of risk vs reward. Let me explain before dismissing my argument.
Do you know why there are so many “Big Pharma”? It has to be big, because most of the smaller ones simply can’t survive for long. Developing drug is a difficult process. Whopping 90% of compounds fail during the three phase clinical trial period, $40k per patient in the trial, hundreds of millions of dollars per trial.
Every failed trial is a signal for career change for people who develops drugs. You think people will invest in such risky business without expecting a big reward? Asking for low drug price without lowering the clinical trial cost is assuming drugmakers can print money. Unfortunately biology doesn’t work like that.
A lot of us who develop drug are doing it because we believe this is a meaningful thing we can do in the world. They are just hidden behind investors like Shkrelli, who isn’t even anywhere near the actual people who make the drug. But we do need the money to develop the drug. A lot of it.
I hope you understand.
The real solution is to lower the cost of the trial. And I maintain that the drug company isn’t in control of this.
the simple math is these companies spend less than 20% of their revenue on drug development and have operating margins over 40% so prices don't need to be as high as they are
You need to take into account of the fact that patent clock starts ticking as soon as you file it, whether the drug gets approved by the FDA or not. The drug development process itself eats away that exclusivity at about 12 years on average, so that 40% needs to account for the time when you don’t get another drug approved (remember the 90% failure rate).
Drug business is EXTREMELY hard, and investors aren’t running a charity. There needs to be high profit margin to motivate people who fund the discovery and development process for new drugs to come out.
As I keep saying, once we bring down the cost of clinical trial (which is totally artificial, and not under the influence of pharma), you will bring down the risk and cost of drug.
Again, you are pointing the finger at the effect of high medical cost rather than the cause. Ask yourself: why are doctors in US paid 3x the amount compared to UK? Do they raise price because of the drugs? No.
Well, if producing a working, custom disposable web application doing exactly what I need to do in half an hour for data processing is bullshit, I don’t know what isn’t.
The point about using the term bullshit (as used by Frankfurt [1]) is it implies indifference to truth.
So the paper is pointing out that, rather than thinking of these models as being mostly truthful, but which sometimes hallucinate, it's more accurate to think of them as bullshit all the time.
It doesn't mean that the bullshit is wrong (although it could be). The model itself doesn't care.
I think a weakness of the paper is it doesn't give a simple definition of what it mean by bullshit. It says "bullshitting, in the Frankfurtian sense" which leaves you to perhaps read his book which can probably be interpreted various ways.
If you look at the wikipedia page for On Bullshit it has:
>Frankfurt determines that bullshit is speech intended to persuade without regard for truth.
Based on that ChatGPT isn't really bullshit. It doesn't really have intent to persuade. It's more like a souped up search engine that takes in a bunch of documents and tries to process the data in them in a way which matches your query. The truth or not out is largely determined by the truth or not of the data going in. And that seems rather good with ChatGPT.
The add glue to Pizza stuff was a different LLM with different input data. It had read a reddit thread where someone had jokingly suggested that. A challenge with these models is to filter bullshit out of the input data though that is an issue with humans too.
Why not make lightbulbs that produce the 350-400nm wavelength to stimulate/prevent eye elongation? I’m sure companies like Philips would love to sell you therapeutic LED bulbs at premium.
It’s probably the brightness, too. Outdoors during the day under a big tree is brighter than basically any indoor lighting you see outside a movie set.
Brighter, and not by a small amount. It's not entirely visible light, but insolation is roughly 1000 W per square meter at noon. I'm not sure I have 1000 W of bulbs in my entire house.
I read an interesting article by someone that set up "daylight" lighting in their home by essentially tiling their entire ceiling in high-efficiency LED panels. He offset the electricity costs using solar power and batteries.
The gist of it was that it very markedly improved the mood and general wellbeing of his entire family, and visitors would often comment about how "nice" his house was without being able to explain precisely why they felt that way.
Conversely, imaging living and working in a dark cave.
No one:
IT infra: let’s move our repo platform to SaaS!