If you are using AI like you use Netflix (something you like but can live without if it gets too expensive) then you have a point. But if you are relying on AI for your business, the cost of AI sooner or later is going to be passed on to you and it's going to impact your margins. And it's probably goong to be sooner, rather than later.
The amount of money these AI companies are burning is unprecedented and there is simply not enough money in the economy to keep it going at a loss for ten years. Google (Google!) went cash flow negative and is issuing bonds, basically asking for a loan. That's the reason markets are getting so nervous.
And yes, choice is good but how many of these new models are trained from scratch vs distilled from frontier models? If OpenAI and Anthropic go down in flames how much of the cheaper choices we see today are going to keep evolving? I am not trying to make a prediction either way, I am just pointing out the uncertainty. Which, again, is fine if AI is a commodity you can easily cut, but not great if your company is betting big on AI.
It's stated in many places that inference is profitable (margin 70% to 90%), only research is expensive. Hyperscalers are burning through their cashflow training new models while inference-only providers are printing money.
If all the research went away tomorrow, people are still going to sell inference at current prices or higher. There are already open weights competitive with proprietary models. There will be no lost capability.
If businesses find inference useful today then it doesn't have to drastically improve in a short timeframe anymore. History is full of inventions which became "good enough" and didn't improve much or at all for a long time.
eg: Western society runs on radial tyres which have seen only marginal improvements for the last 50 years.
(yes there have been some small improvements in compounds, tread patterns, TPMS, etc. hardly drastic revolutionary changes to the tyre industry)
> It's stated in many places that inference is profitable (margin 70% to 90%), only research is expensive.
It's been stated by the AI companies, which are known to routinely lie to our faces, and have a vested financial interest in making people believe they will be able to turn a profit at current prices. In other words, it's one of the most unbelievable claims out there, and you shouldn't believe it for a second.
Even assuming that inference is really profitable today's models don't learn new things by themselves, except in the limited sense of temporarily storing everything they need for a conversation in their context (and maybe leaving themselves little notes in md files like the guy from "Memento").
The difference with tyres is that If today's LLMs had been invented and had become "good enough" 50 years ago, you would have a cutover in their knowledge that excludes 50 years of information. Every "write me a program in Rust" conversation would involve LLMs filling up their context trying to learn Rust programming from scratch every time and probably doing a very bad job.
An example of that was when Fable disproved that mathematical conjecture and HN was full of other people who fed that information to other models (or Fable itself) and received incredulous answers from their LLMs.
If something is proven true or false in math, the world of science moves on and that new piece of information can be used to build, prove or disprove other things. But an LLM is excluded from learning even from the very thing it just helped demonstrate and needs to re-discover it over and over again. In order to have an LLM that "lives" in a world where the Jacobian conjecture is false, you need to train a new model and add that information.
Yes, it's like saying "we took off a big chunk of his brain but look! He can still breathe autonomously, swallow food and walk almost straight, which is like 95% of what he did before!"
At Amazon, due to their toxic work culture which led to high turnover, the joke was Amazon will run out of new hires, it will have to rehire previous employees.
But I want there to be an incentive to develop smaller, private models that can run on local machines that I own. I want that business model to win.
If it costs me slightly more in the short term, but I don't depend on any providers' price gouging as soon as I fully depend on them, I will consider it a win.
From COVID-era discussions (when virologists were briefly the stars of every talk show) I remember one explaining that it was less about fatality rates per se and more about the length of time you could carry the virus around and be nearly asymptomatic while still able to infect others.
I understand the jury is still out on whether a virus can be considered "alive" but, like us, it is capable of replicating itself and mutating. In that sense, it benefits from the same evolution strategies as more complex beings: a strain that gets its host very sick very quickly gets a lower chance to spread to a new host and multiply.
This creates an evolutionary advantage for strains of that virus that are less aggressive or at least develop the worst symptoms more slowly and more covertly.
Yeah. HIV is a good example of this. Without treatment, it is deadly pretty much 100% of the time. However, it takes a long time after the shut down of the immune system before a systematic infection takes over and kills you.
That allowed for a deadly disease that's somewhat hard to spread (mostly just through sex) to ultimately go on a rampage.
So without concern for the humans with HIV* there an argument to be made that treating symptoms without curing made it spread more?
*obviously, this is just hypothetical. It’s important to care about the life of those with HIV. No banish them all to something like a leper-colony. Although it explains the logic for those at the time they existed better than a religious one did.
HIV specifically targets the immune system. There's no way to just treat the symptoms.
The treatments we have now also decrease the risk of spread significantly.
It's a bit like the chickenpox. Once infected, you always have chickenpox ready to burst out in the future as shingles. But for the most part, it's dormant and you aren't infectious.
HIV treatment does the same. It doesn't clear your body of HIV, but it does decrease the HIV load to such low levels that it can be undetectable. That, in turn, decreases the likelihood you'll spread it.
HIV has only really been known for ~40 years. And for at least 10 to 15 of those years research into treatment was limited and stigmatized as it was considered a "gay disease".
The modern treatment regime was developed around 2010. That is, about 15 years.
I'd argue that with the timeline of the disease that's not recent. What's become more recent is the mass availability of treatment and the significantly reduced cost of treatment.
There's a significant difference in the risk and mechanisms of transmission between these two viruses. The modern effective reproductive number for HIV is less than 1, about 30-50% lower than ebola in similar subsaharan countries [1][2]. You can get ebola from contact with an infected person's sweat during the active phase [3]. You can get it from their semen years after they have apparently recovered [4]. We don't have treatments that work in this longer term. The drugs that we do have are used during the active infection period to reduce the probability of death during the crisis. Folks are working on reducing the longer term infectiousness but it's a ways off yet [5]. We also don't have pre- or post-exposure prophylactic treatments for the medical workers who are at the highest risk of infection or the family members--the most common transmission mechanisms for ebola are home caregiving and contact with traditional burial practices. In this context of containing an active outbreak, quarantine is mostly helpful in reducing pressure on the medical system for a short term.
Compare this with HIV, which can be rendered untransmittable with modern treatments, which is primarily a sexually transmitted disease, which has pre- and post-exposure treatments. It's simply not very efficient/effective to exile millions of people with a lifelong latent infection and little risk of transmission.
The instinct towards ostracism of those who are perceived as unclean is some pretty primordial lizard brain shit which was a great rule of thumb two thousand years ago, along with wearing garments made of only one kind of material. It's actually actively harmful to the process of stopping infection. It leads to fear, distrust, and reduced reporting, hindering the medical system's ability to reach the people who most need to be reached, and encouraging the spread of superstition and suspicion of pre- and post-exposure treatments. In both diseases, the actual infection risk is modest compared to an airborne virus like COVID.
>I understand the jury is still out on whether a virus can be considered "alive"
I remember way back in med school in the mid-70s our infectious disease professor asking this same question, in a philosophical as much as a mechanistic sense.
Here is the fun thing to think about: If viruses aren't alive, neither are men. Men, too, lack the biological means of reproduction. On the other hand, parthenogenesis may even rarely happen in female humans.
Where are all the apps?
It's mostly visible in AI tooling itself. Harnesses, vibe coding tools and stuff with "claw" in the name saw a cambrian explosion.
And maybe using AI to use AI better is just masturbatory. But coders want interesting problems to solve. Pros also need software ideas they can monetize. And what problem is attracting more investment in money, time and neurons than the problem of making AI productive? (I am referring only to problems that can be solved in software....)
So the thing with AI is that right now it is both a tool AND a potentially very valuable problem to solve, that's why most of the AI "productivity" gains go into AI itself.
At one point this self-refetential phase will have to end and people are going to see if these new AI tools, harnesses.claw-things are actually applicable to things people are willing to pay the real prices for (not the subsidized ones).
I find the music example very illuminating, thanks!
Looking into US Copyright for songs there are two different kinds:
- one for the composition, the musical idea, music, lyrics.
-one for the recording, the music taking shape in a format that someone can listen to
I don't think this is how software licenses work, as they cover the code itself, rather than the ideas (the specific recording rather than the composition, in the music example), but it's an interesting way to frame why using LLM this way is, if not illegal, at least unethical.
I'm not saying that CEOs (or devs, for that matter) lie. But on AI I don't think we can rely on any self-reported results, positive or negative, based on surveys.
There is just too much incentive to say... no, to BELIEVE... both that AI yields 10x productivity that AI is useless.
I am swinging wildly between the two too, personally. The more time I spend with AI, the more I am developing this split personality where one part of me says "I hope this thing blows up before I lose my job and my children never have the chance to have an office job again" and the other one says "AI is actually not easy! You have to know how to use it well, deveop tools, plan, curate your context... This means I am acquiring useful skills here, tring to port Flappy Bird to COBOL".
And obviously, depending which side controls my cortex in that moment, I may err on the "AI is useless crap" or the "AI all the things!" side
I think an interesting analogy for what many of us are experiencing here is the phenomena of Doom Scrolling; deep down we know we should put it down (and go outside), but the immediate experience of it and the value it feels like it’s offering in the moment has you keep scrolling and scrolling.
Similarly many have reported a sense of say programming productivity but a more objective reflection later on reveals the myriad issues with constantly and subtly heralding in large quantities of lower quality code and blowing past any caution or rigourkus discipline that would come with the laying down of lines of code “by hand”.
I have also decided to do this as soon as the burden of lying about my AI usage becomes too onerous. Right now at Cisco, there are no mandates, only very strong recommendations with the explicit threat of being "left behind" if you fail to comply. Some teams have included AI usage in their personal KPIs which affect bonuses and promotions, but mine fortunately has not.
Once the execs or my personal manager implement AI requirements, I'll have to start lying, which I really prefer not to do. If they start tracking, then I'll have to vibecode a script to make bullshit requests to the API each day. And if they start auditing, then I'll just check out (more than usual) and wait to be fired. They're only hurting themselves with this shit.
The amount of money these AI companies are burning is unprecedented and there is simply not enough money in the economy to keep it going at a loss for ten years. Google (Google!) went cash flow negative and is issuing bonds, basically asking for a loan. That's the reason markets are getting so nervous.
And yes, choice is good but how many of these new models are trained from scratch vs distilled from frontier models? If OpenAI and Anthropic go down in flames how much of the cheaper choices we see today are going to keep evolving? I am not trying to make a prediction either way, I am just pointing out the uncertainty. Which, again, is fine if AI is a commodity you can easily cut, but not great if your company is betting big on AI.