Hacker News

No Bailout for the AI Bubble, Anthropic IPO Protesters Say

50 points by cdrnsf ago | 34 comments

ethnrob |next [-]

> We have AI deciding medical outcomes and treatments in this country. Shame. Shame.

What about cutting triage times for strokes, solving novel protein structures, or catching early-stage cancers? I'm not saying AI should be making final judgments on people's health, but this framing feels pretty one-sided.

mhitza |root |parent |next [-]

AI shouldn't be allowed to make decisions. Unless you want a more opaque future in which no corporate/state person is even responsible of negative outcomes "because computer says so".

nancyminusone |root |parent |next |previous [-]

What about it? Should people be socially treated worse even if the medical treatment is better?

apercu |root |parent |previous [-]

Capitalism- it’s not about health, it’s about novel ways to reject insurance claims.

tt24 |root |parent [-]

Are insurance companies supposed to accept all claims without question? I don’t understand your view

unrented7977 |root |parent |next [-]

How do you feel about insurance companies denying and delaying in the hopes the patient dies before they have to pay for treatment?

tt24 |root |parent [-]

Evidence that they’re doing this with the specific intention of the patient dying to avoid having to pay for treatment?

mikestew |root |parent |next |previous [-]

No one is saying that but you, I fail to understand how you got that out of OP's comment.

tt24 |root |parent [-]

What’s wrong with finding novel ways to reject insurance claims?

mikestew |root |parent [-]

I can "just ask questions", too: where did I put that can of Sealion-B-Gone?

tt24 |root |parent [-]

Yup, this is about the level of discourse I expected from someone who hasn’t thought out their positions and just goes by vague notions and gut feelings. No surprises here ;)

mindslight |root |parent |previous [-]

Valid claims for services performed by a licensed doctor on an insured patient, yes.

If this is too much there is a simple way to avoid it, the way most everybody else does - don't write insurance policies that purport to cover people's medical issues.

tt24 |root |parent |next [-]

What makes a claim valid vs not? Who determines that?

unrented7977 |root |parent |next [-]

A licensed medical professional who has seen and laid hands on the patient. Not a bureaucrat or an algorithm.

Nobody has any right to make decisions about a patient's healthcare other than their physician.

tt24 |root |parent [-]

A licensed medical professional has determined that 10,000 people require a treatment course that costs 1m per dose and requires 3 doses per day for 10 months. Now what? Is the insurance company obliged to go bankrupt and liquidate to support 10k patients?

Scarcity exists and rationing must occur to resolve that scarcity.

mindslight |root |parent [-]

If said insurance company has agreed to cover 100% of costs? Sure, why not? Make bad bets, go bankrupt. But in the situation you describe, there should be an investigation into fraudulent conveyances. It would seem the insurance company rapidly took on a load of unsustainable business - presumably so they could book the revenue as profit and put it into their personal pockets before the bill came due.

Does your logic apply to houses as well? Pay into home insurance for decades, have a structure fire, then insurance company denies claim "sorry, scarcity exists! not everyone can have a house" ? I'd say that's preposterous - the whole point of insurance is to keep the insured party whole.

tt24 |root |parent |next [-]

> If said insurance company has agreed to cover 100% of costs?

This isn’t the case here. An insurance company doesn’t agree to cover 100% of costs without first reviewing what it’s paying for.

s1artibartfast |root |parent |previous [-]

You are representing insurance companies as if they signed up to pay any and all medical expenses, which they absolutely havent!

If I carry wildfire insurance and then have a I kitchen fire, I wouldn't expect them to cover it.

Insurance isnt to make/keep you whole, in almost any cases. It might offset very specific issues, or mitigate a fraction of the cost of others. I have $10k life insurance on my infant. Will that make me whole?

As a hypothetical alternative, unconditional approval can be discussed, but has tons of problems around cost control. We might as well all put money in a bank account and race to spend it.

mindslight |root |parent [-]

They signed up to offer health insurance, a product that most everyone is forced to buy, while having very few choices about who they can buy it from. So no, I am not receptive to the idea that when it comes time to provide healthcare, that companies should be able to use fine print to weasel out of actually doing that! As a matter of public policy, the entire point of insurance is to provide coverage for small probability high cost events - expecting a purchaser to weigh which particular small probability events might end important to them runs directly contrary to that.

In general, the overarching dynamic of our system is the worst of both worlds - central planning for the provision of healthcare capacity, but then free market for extracting wealth based on the resulting shortages. I'm not saying pushing the way I'm pushing is a solution, rather merely pointing out the hypocrisy of the extractors to wax poetic about free markets when they're taking money, but then appeal to collective limits to avoid having to spend it.

If you want to talk about actual reform, sure. The easy response to what I'm saying is for health insurance companies to require 10% copays (or whatever purely-dollar-based schedule provides sufficient financial incentive for a patient to keep costs down). But of course that would require transparent pricing in a functioning market, rather than the state of shameless collusion and fraud the industry has entrenched itself in.

Also note getting back to the original topic, the reason people are so universally against AI writing justifications to deny healthcare is precisely because it is not a positive sum open market, but rather a negative sum closed market. Any "innovation" here is necessarily based on making things worse for actual patients the system is meant to serve. If you want AI in healthcare to be positively received, then you need to address this dynamic first!

s1artibartfast |root |parent [-]

Seems like you are starting from a distorted view of reality that is hard to build any productive change on. We aren't living in the same world.

The companies dont think they are selling a golden ticket to unrestricted care. Buyers dont think they are getting an unrestricted ticket.

You may want it to be the case, but asserting that is the contract everyone signed is delusional.

I dont think denials are negative sum. Every ai or human denial is money saved out of my pocket instead of waste on unnecessary care. I want more of them!

I think we agree markets and insurance benefits should be more transparent.

Get rid of policy specific pricing and make it all up front.

Insurance policies should be clear, Eg Generics only or unlimited access. Maybe insurance should dollar denominated like your car or home, eg insured up 100k/yr or 1 million lifetime.

I dont know if care would be better, but it would be more certain.

I agree that the government should stop restricting supply. Either lower accreditation standards or make lower tiers of doctors and HCPs. Let the market sort it out. A doctor with no bs and a 2 year MD should be paid less accordingly.

reverius42 |root |parent |next [-]

> Buyers dont think they are getting an unrestricted ticket.

They definitely do, to the extent that the plan says they will pay for X, if you need it, and your doctor thinks you need X, leads to a reasonable expectation that they should pay for X. But wait, the insurer decided you don't need X! Hooray, more profit for the insurer, and you don't get the care you need (and that you reasonably thought you had an "unrestricted ticket" for, based on need).

mindslight |root |parent |next [-]

> Hooray, more profit for the insurer

nit: This is fallacious. It would be great if the world were so simple. Outright profits are capped as a percentage of revenue (ACA). The real dynamic here is the ever growing bureaucratic complexity that creates and illusion of savings/sanity/control but actually does the completely opposite.

The other play that has/had been happening is the same people owning/controlling insurance companies also owning providers, as a way of getting money out (booked as expenses for the insurance entity) without running afoul of that cap.

s1artibartfast |root |parent |previous [-]

Lwts be simple, do you think that - That you have a golden ticket?

mindslight |root |parent |previous [-]

You're the one invoking an idea of "golden tickets", as if access to healthcare should be some rare privilege rather than a reasonable expectation in a functioning society. And I am well aware I'm talking about how things should be rather than how they are.

> Every ai or human denial is money saved out of my pocket instead of waste on unnecessary care. I want more of them!

If you believe this universally, you can simply just cancel your health insurance and you will have solved the entire problem as you see it! But presumably you will only hold this position until you yourself end up needing substantial care.

Also from my experience it's not even outright denial is the problem, but rather the soft denial that a patient/family (already having to deal with the actual medical issues!) then needs to spend an extreme amount of time and effort to reverse. Never mind the overhead of the staff at providers whose whole job is to navigate this system and write justifications for the insurance companies - a core of rot the system is drowning in.

> Get rid of policy specific pricing and make it all up front.

Yes, I elaborated on some other things to show you we actually probably agree on a lot. I'd say our disagreement here is based on you having internalized a negative sum justification of the medical industry - one that absolves the insurance companies of responsibility for the position they've chiefly created and now sit.

On the whole I'd say providers have done as much, if not more, to corrupt the system. I can rant for hours about their routine-fraud-based billing practices, which tbh I have suffered more harm from than from insurance shenanigans. But here we're talking about insurance companies, and I don't think it's right to let them escape blame for their part - specifically the ever-growing-complexity paperwork bureaucracy they've eagerly created.

mindslight |root |parent |previous [-]

Sorry, my sentence structure was a bit ambiguous. By "valid" I mean conforming to the two conditions I followed with.

The licensed doctor is the one exercising judgement of whether a given treatment is medically prudent. An insurance company asserting that it is entitled to create its own medical opinion on which to deny care is sophistry. If the insurance company has a reason to believe that the doctor is violating their professional responsibilities, then the insurance company should take it up with the medical board.

(I think our healthcare system needs much more wider sweeping reform than this, although I'm not predisposed towards single payer. My angle here is pointing out the corruption of the current system)

s1artibartfast |root |parent |previous [-]

They dont! Thats the point of claim denials - they dont cover it

joegibbs |next |previous [-]

Is there any evidence at all that they will need a bailout, or that the government has considered giving them one?

|root |parent |next [-]

whateveracct |root |parent |previous [-]

BAAJA is getting rammed through by a lame duck Congress, conveniently timed with the IPO..

boshalfoshal |next |previous [-]

How are random protestors so easily able to identify a complex financial and economic bubble? Feels awfully presumptuous to claim this is a bubble already. It really does feel like two completely different worlds between being in a high tech company/area and the average American. These appear to be way too kneejerky and not actually attacking actual problems that powerful AI might lead to. Also what bailout are they referring to?

To me, the utility promises of AI have basically been met. This technology basically obviates verticals of knowledge work that can be done at a computer, imo. Theres clearly utility here and the technology is not a bubble. Perhaps certain companies will go bust in the process, but pandoras box has been opened. There are far more pressing things to argue against than a "financial bubble" and "datacenter water usage," but I think these issues are way larger and more complex than a group of average people are going to platform.

To paraphrase an anonymous openai employee on twitter: I think people want this to be a bubble because they are not emotionally prepared for the world where it isn't.

thereader12 |root |parent [-]

There is definitely analysis that has been posted to hackernews itself hinting at a bubble. It is interesting how disparate people's ai experiences are, cause the promises haven't all been met from my pov

millspore |next |previous [-]

[flagged]

tt24 |next |previous [-]

Have they asked for a bailout?

carefree-bob |next |previous [-]

What is the basis for demanding that Anthropic remain a privately held company, and, uh, how would one go about preventing an owner of Anthropic from selling their stake to a willing buyer?

Seems like a really strange thing to protest, especially since I remember people being angry at public companies going private, but now I guess they are angry at the reverse.

tancop |root |parent [-]

If Anthropic gets included in the Nasdaq 100 and they go bankrupt, the government will have to bail them out with tax money or let everyones portfolio go into free fall. Other tech stocks are correlated because of AI investments so they will crash together.

The other risk is circular funding where debt backed by Anthropic securities can enter default. We might see a chain of bankruptcies affecting the whole market.

999900000999 |previous [-]

Here’s what’s going to happen. They’ll keep expanding and over extending.

Even when it becomes mathematically impossible to get a decent ROI. Then when it falls apart they’ll be declared a national security priority and get billions upon billions in tax payer money.

You don’t want us to fall behind the COMMUNISTS do you. Thus to beat the communists we must become them. How much of Intel does the government own at this point ?