That's true. Good Clinical Practice guidelines say that Institutional Review Boards must evaluate compensation is not too high otherwise it constitutes "undue influence".
...but I think that's a terrible idea. I understand that, to come up with a far-fetched example, high compensation could result in some weird edge cases where, for example, a Phase I trial with healthy volunteers and a non-preventive drug candidate (e.g. an antibiotic rather than a COVID vaccine), but you're confident that drug is going to kill 99% of the volunteers so you offer $10 million to each volunteer. A totally informed volunteer might view this as something akin to a life insurance policy that pays out to the family even upon suicide. I don't personally see an ethical problem with that so long as volunteers are really well-informed, but I could see how others and the law would be uncomfortable with the ethics of such a situation.
But aside from that far-fetched situation, it seems harmful to limit compensation... especially if it's true that patient compensation is a small part of overall cost. Perhaps if every trial doubled compensation to 10% of today's overall costs, absolute overall costs would decrease because less effort has to go into patient recruitment.
I do see one real-world issue with high compensation, but it has nothing to do with ethics. It has to do with statistical interpretation. If you start dangling too much money to participate in a trial, you may change the kind of people who participate in the trial. These individuals might not be representative of the actual users of the drug. The problem is similar to running a drug trial in the US versus Japan. The US has a pretty diverse population with respect to genetics. Japan, in contrast, is almost entirely ethnic Japanese. If you were to somehow succeed in using Japanese trial data to market an acne cream in the US, the same drug that cleared up Japanese faces might result in Venezuelans having their noses fall off!
However, I'm cautiously optimistic that higher compensation could make trials more representative. In my mind, I could see how African-Americans, who tend to participate in trials at a lower rate than their Caucasian fellow humans, would, if only for the shameful history of the Tuskegee Syphilis Study, require greater compensation to take the risk of participating in a trial.
I would think the worst of such edge cases should be prohibited by law anyway. A set of more subjects is presumably a superset of a set with less subjects. So any issues can be addressed by screening them better. I suppose there could be a greater problem with fraud as people claim to have issues when they don't.
It just seems unfair to subjects to me. Not to mention paternalistic. Every other actor in the process gets enriched based on (limitless it seems) market rates for their contribution, except for the subjects who take on all the risk.
As you said, at the end of the day it's an ethical guideline that must be upheld, more than something based in economics. We do compensate participants as well as we're allowed within these guidelines!
I agree with your points, but compensation must be approved by the independent institutional review board, who is accountable to national ethical review boards and thus acts very conservatively.
I worked at a small-to-mid-size clinical research organization for a hot minute. That company had its own set of - let's say "unique" - problems*, but the author touched on many of the issues I saw in my short experience at a CRO. I was in an odd (for me), pseudo-managerial, HR-type role at the company (a role in which I'll be the first to admit I wasn't good at) so perhaps I can offer some additional perspective.
But first, I want to give some background on an important part of the CRO world: the "clinical research associate" or "CRA". CRAs are are the CRO's meat-and-potatoes, boots-on-the-ground, core, etc. after contracts have been signed. A CRA is the person who actually goes to clinical trial "sites" (e.g. a doctor's office, a hospital, or, even a dedicated facility for clinical trial participants). The CRA goes to each trial site, talks to the doc, makes sure participant recruitment is going well, checks/writes documentation, and follows up as the trial goes on. In other words, the performance of the CRAs strongly & directly affects the performance of the clinical trial.
Anyway, my top issues with clinical trials today are:
(1) Poor patient recruitment - this is a major problem. There is, of course, lots of variability between trials and CROs with respect to recruitment, but it should be telling that "trial rescue" is often done (at least where I worked) as a result of the incumbent CRO's inability to recruit enough patients.
(2) (Related) Limiting pay to study participants: although it's up to the institutional review board's interpretation (IRB, an ostensibly independent entity that's supposed to sign off on the trial's ethics), good clinical practice (GCP) guidelines states that paying participants too much would be "undue influence". I have no idea how this plays out in practice, but I think it's a bad guideline. Most of the other parts of GCP are good things - stuff like informed consent and not coercing subjects - but this one is (again, in my opinion) probably raising trial costs through recruitment delays or total non-recruitment, even accounting for the extra cost of paying participants more.
(3) Clinical Research Associate (CRA) training & hiring: this isn't unique to the CRO industry but it seems like a bigger problem than in other industries. I've seen US median salary ranges from $70,000 to $120,00 from publicly available sources but my time in the industry rarely demonstrated numbers that high. Although CRAs don't treat patients directly, it is a role that requires medical knowledge, attention to detailed paperwork, and good communication skills. There seemed to be a lot of cutting corners on training & hiring CRAs - e.g. hiring highly experienced CRAs from India and putting them into entry-level roles in the US on the cheap. I don't mean to say there aren't great experienced CRAs from India who are also great when dropped into an entry-level CRA role but doing so as a matter of unofficial policy is a bad idea: India's drug industry has a different regulatory environment (although there is significant cooperation between FDA and Indian manufacturing facilities) and, perhaps more importantly, there will be a wide variation in communication skills simply because english is a second language (or, in the case of India in particular, "Indian english" has taken on a life of its own).
(4) Over-promising, under-delivering - true of any contractor even outside the CRA industry. Even when working on a fixed price basis, there are perpetual "change orders" and arguments over change orders.
Side-note: if you want to learn more about the process by which drugs come to market, I strongly recommend New Drugs by Lawrence T. Friedhoff. It's short and the HN crowd won't have any problem understanding. Amazon link: https://www.amazon.com/New-Drugs-Insiders-Scientists-Investo...
* "Unique" problems in the sense that they were personality-based. I won't go into specifics nor do I wish to speak negatively of my prior employer who, to their credit, took a chance on me that just happened to not work out. But I will give some general advice: if, during a job interview or in early conversations, senior management insults his/her own family and dumpsters their own employees, you're going to have a bad time.
An interesting company I talked to a while back working on (1) in the spaced is deep6.ai. Interesting approach and there is a lot of work that can be done here sourcing patients more effectively. I think they said something like only 10% of studies are able to be carried out due to lack of patient recruitment.
I would add to your list of problems: Decades of easy-money milk-the-cow style companies (Often CROs) with very outdated systems with folks sittings in cheap countries to do some data change requests and manual tests. Truely horrible and outdated tech, mixed with bad salaries, high turnover and bad management.
I used to work for a company were because of timezone issues a wrong date was entered, CRA requested correction (edit manually, sign, scan) and wrong assumptions etc we had to change 7 times the whole thing and send CRA onsite and recorrect the mess. In short: Inefficient mess.
I left that industry quickly and can say I that others are better. Lab tracking is way better with diagnostic samples with private labs or when mixed with good logistics partners. It is not everywhere as bad as this.
As someone who has actually flown on the Concorde, I can say that I'd gladly endure its tight seats again and empty my wallet in exchange for a shorter flight. Heck, if I had the option between spending $20,000 for a NY-London roundtrip on an Overture (the same price for that route on the Concorde, inflation-adjusted) versus spending $20,000 to charter a long-range private jet (likely a significant underestimate), I'd go for the Overture in most cases.
However, although I'm rooting for any company that's making a sincere (as opposed to fraudulent) attempt at bringing back supersonic travel, the hardest challenges may still be ahead for Boom. The biggest one is the need to find or build a new engine. They've recently redesigned the Overture to use four engines instead of two, which should ease required engine specs, but there's no engine that would meet the reliability, noise, fuel consumption, and dimension requirements for a supersonic passenger aircraft.
Related to the engines: money. It sounds impressive that boom raised at least $150 million, including $60 million from the US Air Force (which has the added advantage of creating a new customer segment in the military)... until you learn engine development alone would require in the ballpark of $6 BILLION of capital. Aviation history is rife with examples of amazing, innovative aircraft designs that failed because no suitable engine was available.
Also, Boom leadership has set some ambitious goals, which makes me a bit skeptical. They plan on using sustainable aviation fuel (SAF). Great! But now they not only need to create a new engine, they need to create a new engine that runs off of a new fuel. Additionally, they've set a goal price of $5,000 for a New York to London roundtrip whereas Concorde would've cost $20,000 for the same route. Heck, I once paid $8,000 for a Boston to Tokyo roundtrip business class flight. Nothing wrong with setting such a goal (and Boom isn't even the party that sets route prices) and it's OK for marketing claims to be a tad optimistic, but this tests the limits of credibility.
Lastly, there's the issue of possible routes, which is primarily limited by noise constraints. Unlike the Concorde, which needed afterburners to produce sufficient thrust for takeoff, Boom is going for a no-afterburner design. While this should expand the number of airports the Overture can use since afterburners won’t be blasting the neighborhood, you’re still not going to be able to fly over land. Boom suggest 500 routes are supersonically viable[1], which I’d assume means “pairs of international airports separated mostly by water”. We might be talking about something like 50 actual airports. only a fraction of those routes are not just supersonically viable, but economically viable. Of course, commercial aircraft are designed for particular types of routes. An Embraer ERJ-145 regional jet and the Boeing 787 long-range wide-body jet fly different routes. I’m not expert on this though; maybe 500 routes is plenty for a “total addressable market” in the aviation industry,
To bring it all together: my big issues with Boom are, one, engine development and, two, the choice of “hard problems” they decided to take on (specifically, SAF & cheap tickets). My hopes are that the engines are in development, using SAF instead of conventional fuel isn’t a big deal if you design for from the start, and the $5,000 thing is more about saying how low, hypothetically, an airline could price tickets while making money. I’d also like to know what the current status of the state-of-the-art is in quiet supersonic flight. NASA’s quiet supersonic demonstrator, the Lockheed Martin X-59 QuSST, combined with regulators’ desire to decide on supersonic overland travel in 2028, would open up new routes like JFK-LAX for planes meeting noise requirements, should regulators decide to allow it.
My hypothesis on Boom’s design choices? Quiet supersonic cruise is still technically challenging and has an uncertain regulatory future, and the political tide may be turning towards greater regulation on fossil fuels. So, by using SAF, Boom ensures that their plan will at least fly in an uncertain regulatory future, even if there’s no overland flight. And, using what they learned developing the Overture, they’ll be in a position to develop a quiet supersonic transport should regulators give the green light.
[1] I’d interpret routes to be something like airport-pairs, as in Laguardia-Heathrow would be one route. If you Boom could fly from three airports in the US to or from three airports in Europe, you’d have nine routes (3*3). This article talks a bit about the lack of clarity with Boom’s “route” number: <https://leehamnews.com/2021/06/04/hotr-500-destinations-for-...>
I bought an MX Master 3 for Mac mouse for my Silicon Mac Mini about a year and a half ago but quickly shelved it because the cursor would frequently make irritating little jumps. I replaced it with an Apple Magic Touchpad (I don't love the Magic Mouse) because, at least at the time, I was under the impression that the jumpiness was due to a OS-level bug affecting all non-Apple mice. I did have Airpods at the time but I don't recall if I ever attempted to use them alongside the Logitech mouse.
The timing is fortuitous because, on a whim last night, I got my MX Master 3 out of storage and put it on a charger. I figured that, with ~18 months of macOS updates, perhaps it'll work with my Mac Mini on bluetooth or using the Logi Bolt USB adapter, which I don't own yet (I guess the Bolt is different from the Unifying Receiver... although irritatingly neither comes in USB-C).
I'm intrigued by the hypothesis that Logitech is remaining quiet because their products may be violating FCC rules, specifically, part 15 of the FCC rules: "this device may not cause harmful interference". If I can replicate the AirPods issue, or if I see some other weird behavior with the Logitech mouse, I might just break out my HackRF, YARD Stick One, or Ubertooth One. I'm a total noob at radiofrequency stuff, and I'm only slightly less of a noob with running specialty software, but I wonder if I can spot some unexpected RF emissions.
> specifically, part 15 of the FCC rules: "this device may not cause harmful interference".
I think you misunderstand what harmful interference means - it has a specific regulatory meaning - and the fact that it causes you grief with your other Part 15 devices doesn’t apply. If that were the case the FCC would be spending all their time just dealing with everyone’s shitty stereo equipment not working.
> but I wonder if I can spot some unexpected RF emissions.
I highly doubt it - this stuff was already certified. On the other hand it’s not surprising that devices physically near running in the same frequency band will have some undesired operation - this is why safety critical communications don’t operate in an unlicensed band. Spread spectrum isn’t completely magic (as an EE I’m still sometimes amazed any of this shit works at all) —- or maybe I should say it is magic but it still has to come to terms with physical realities.
Logitech Unifying and Logi Bolt are incompatible. Both them, classic Bluetooth, Bluetooth Low Energy and almost all other input-devices with dongles all use the 2.4 GHz band, and all these types of devices can interfere with one-another — but there are no rules against that!
Logi Bolt is actually a proprietary layer on top of Bluetooth Low Energy.
Logitech claims that it should suffer less in face of congestion on the band compared to regular BLE, but they have not revealed how: I wonder if they make it worse for other devices by transmitting on more BLE channels.
While I agree with much of your first couple sentences and resisting the emotional urge to think recovery is impossible, you're walking a hazardous line between believing recession timing is effectively impossible and trying to ballpark near term recession risk.
I preach this: nobody can reliably predict recessions but everyone should know recessions will happen. It would be damn near impossible to predict recessions if you had a single large country only trading with itself. Throw in the billions of people in the global economy, each a market participant, and the task becomes intractable. The macroeconomists and policy wonks sound much like you do: identifying real risks and evaluating sentiment in an attempt to convince others, and themselves, that they have special insight. I try not to be a cynic - there's enough negativity on HN - but I really think recession prediction is an exercise in futility.
Still, it's critical to expect recessions. Buying stock in a company that's selling for 100x revenue and would go bankrupt in the event of a recession is gambling. Likewise, leaving the market and putting your money in gold, Bitcoin, your mattress, etc because you think recession is around the corner is likely to burn you. Yes, Michael Burry and The Big Shorters made a ton of money betting on what would coincide with a recession... but some of them would've gone broke had the timing stretched on even longer.
In short, your best bet if you're scared of recessions is the same as your best bet if you're in a boom or in the middle of a recession: buy undervalued assets and hang on to them long enough for everyone else to recognize their value. And, since recessions happen regularly at an irregular interval, your job will be made much easier if those assets can simply survive a recession.
Scuttleblurb, David Kim's SubStack, sends out in-depth write ups of companies. It's technically an investing newsletter, but it's really more about the mechanics of a given industry and how a particular company or group of companies operates within that industry. Similarly, and these aren't newsletters but they do come out at newsletter-like frequencies: the Company Man and the Wendover Productions YouTube channels.
Granted, these aren't specialized in terms of a single industry or single tech field, but, if you're a business enthusiast / love learning about industries you previously didn't know about, these are excellent.
Correct. Aatsmyles was mistaken in relating the goodwill he/she described to accounting goodwill. In other words, aatsmyles shouldn't have used the phrase "account for" goodwill.
Interestingly (at least to a weird human like me), there is something of a relationship between accounting goodwill and goodwill like the value of a brand. The reason why goodwill only shows up on a balance sheet after an acquisition is, I imagine, to follow the accounting principal of conservatism.
Let's spice things up with a hypothetical. I'm going to make up some numbers here so don't go around telling people I revealed some privileged information on HN.
Say you're Mr. McIlhenny, the() owner of a major private company called the McIlhenny Company. The McIlhenny Company's primary endeavor is selling a beloved hot sauce called Tabasco. On the income statement side, McIlhenny has revenues of $200 million and profits of $30 million. On the balance sheet, Tabasco has no liabilities (no long term debt, no payables, etc) and its only asset is cash, of which it has $1 million. Since (equity) = (assets) - (liabilities), this company has a "book value" of $1 million. You might notice that the book value seems absurd - a company that makes tens of millions of dollars a year and has a product with a major following would be a total steal of a purchase at $1 million!
A few purchasers attempt to woo you, and they each make offers for about $100 million. You go with Carl Icahn's offer. Now, the company's book value is $100 million (the balance sheet has $1 million in cash and $99 million in goodwill on it).
Clearly, the day before the acquisition, the company had roughly $99 million in "real" or "intrinsic" goodwill, but that didn't show up in the balance sheet. Why's that? One reason is conservatism. For many intangible assets, there's an art to choosing a number. If you let CEOs put in a goodwill number, many would probably throw in huge numbers as they vastly hype up the value of their brand and reputation. So, instead, we have a market approach to calculating goodwill by using transactions.
However, accounting principals allow companies downwards. So the CEO of Nikola isn't allowed to turn his bogus claims into dollars on the goodwill line item, but he is allowed to reduce goodwill if he buys a battery manufacturer that turns out to be a fraud too.
() It's owned by the McIlhenny family, but no need to complicate things
Shouldn't there be a way to only search multiple forums through some userscript? Only found out about userscript after messing around with an iOS app called HyperWeb.
At least in my limited understanding of userscript's capabilities, you'd still need to explicitly specify the forums you want to search, but I imagine that would solve lots of use cases for search filters.
Google has a limit to how long searches can be. So if it’s a complex one, it can’t be too long.
I think you can only specify searching one site at a time. With site: search filter.
—
Google’s custom search engine’s where you can give a whitelisted amount of sites isn’t very good. At least not any more. It didn’t include lots of pages it’s main index does. Or the rankings were all out of sorts rendering not useful.
To reductio ad absurdum that - every time you see any marketing, eg “Drink Coca-Cola because it’s refreshing”, you should hear “Drink Coca-Cola because it’ll make us money.”
To your point though: “democratize x” makes my eyes roll. It’s overused hip marketing speak.
That's accurate though; coca cola don't care about whether you're refreshed unless it's profitable to them. The refreshment is the means towards the end of profit.
Indeed, contracts can be invalidated if they're abusive. For example, there's a doctrine in contract law called unconscionability. If you buy an airline ticket to Boston, the captain can't say upon landing "Welcome to Boston! You can pay $10,000 to deplane or we'll take you back to where you departed." Or let's say that, when you agree to the terms for a e-commerce site, buried deep in the terms is a clause saying the site gets half the money in your bank account. Sure, both parties might've technically agreed to the contract, but the contract probably isn't enforceable.
*Not a lawyer either, but I've read some introductory contract law stuff. All I know is this: you don't want to rely on the courts to save you, since they generally err on the side of the literal contract, but I don't spend much time worrying about accidentally agreeing to a contract that will sell myself into slavery.
...but I think that's a terrible idea. I understand that, to come up with a far-fetched example, high compensation could result in some weird edge cases where, for example, a Phase I trial with healthy volunteers and a non-preventive drug candidate (e.g. an antibiotic rather than a COVID vaccine), but you're confident that drug is going to kill 99% of the volunteers so you offer $10 million to each volunteer. A totally informed volunteer might view this as something akin to a life insurance policy that pays out to the family even upon suicide. I don't personally see an ethical problem with that so long as volunteers are really well-informed, but I could see how others and the law would be uncomfortable with the ethics of such a situation.
But aside from that far-fetched situation, it seems harmful to limit compensation... especially if it's true that patient compensation is a small part of overall cost. Perhaps if every trial doubled compensation to 10% of today's overall costs, absolute overall costs would decrease because less effort has to go into patient recruitment.
I do see one real-world issue with high compensation, but it has nothing to do with ethics. It has to do with statistical interpretation. If you start dangling too much money to participate in a trial, you may change the kind of people who participate in the trial. These individuals might not be representative of the actual users of the drug. The problem is similar to running a drug trial in the US versus Japan. The US has a pretty diverse population with respect to genetics. Japan, in contrast, is almost entirely ethnic Japanese. If you were to somehow succeed in using Japanese trial data to market an acne cream in the US, the same drug that cleared up Japanese faces might result in Venezuelans having their noses fall off!
However, I'm cautiously optimistic that higher compensation could make trials more representative. In my mind, I could see how African-Americans, who tend to participate in trials at a lower rate than their Caucasian fellow humans, would, if only for the shameful history of the Tuskegee Syphilis Study, require greater compensation to take the risk of participating in a trial.