* From today through June 22, Fable 5 is included on Pro, Max, Team, and seat-based Enterprise plans at no extra cost.
* On June 23, we’ll remove Fable 5 from those plans. Using it after that will require usage credits. If capacity allows, we’ll extend the included window.
* After this point—when sufficient capacity allows us to do so—we aim to restore Fable 5 as a standard part of subscription plans. We intend to do this as quickly as we can.
The "offer, then remove" aspect is a bit eyebrow-raising -- it feels like they are trying to get subscribers to switch to usage-based billing, which makes me wonder if we'll ever get it after that June 22nd window.
If they didn't announce it, you guys would be complaining about slowed progress.
If they didn't release it, you guys would be complaining about fake promises and marketing.
If they released it without limits, the complaints would be about slow responses and outages.
If they didn't add to susbcription plans, the complaints would be about phasing out subscriptions.
If they added to subscriptions with cost reflecting their resource availability, the complaints would be about how quickly it eats limits.
So they choose the middle ground of providing some initial access and assessing if they can satisfy demand, only to still be ignored and accused of trying to get users hooked?
We've already seen that they don't have enough compute, thus the deals with SpaceX for their GPUs. It's very reasonable that they just don't have the capacity to support the subscription userbase on this model.
I can recognize so much of the GPT/Codex generated code long after it gets merged (not by me).
Additionally, the time spent on every agent turn on GPT 5.5 is much longer compared to Claude Opus 4.8, which means iterating on the code takes a lot more patience, and there's a lot more nitpicks to pick when actually using GPT 5.5 to do software engineering.
Feels like GPT-style models are more geared on doing one-shot software vibing (and handling the vibe coded mixture) compared to Claude's focus on actual software maintenance. I got a GPT Pro sub for free and wanted to cancel my Claude subscription so much, but I still keep reaching Claude models a lot more. Frustrating.
this is the line I keep in Agents.md that helps me prevent Codex from playing smart
When a "person" that you don't view as a "real" person repeatedly does exactly what you just told it not to do (often amid false assurances it understands and will avoid doing so in the future), most people get angry.
Compare it to how the kind of people who treat children like property treat their kids, or other examples of keeping people as property.
We were reviewing reports of situations where the models failed to follow directions and there was a common thread of some where when the operator got the model to acknowledge the rule breach, it quoted back something that included swearing.
I don’t have the data to truely look into it, but I did give the instruction to my engineers to avoid it as a “might be a problem”.
But I avoid unnecessary emotion in my prompts because I don't want potentially distracting activations. Kind of like communicating with humans.
> impolite prompts consistently outperformed polite ones, with accuracy ranging from 80.8% for Very Polite prompts to 84.8% for Very Rude prompts.
Unless the mechanism is understood, my assumption is that this is a moving target.
https://www.anthropic.com/research/emotion-concepts-function
How so? Plenty of swearing in lots of training data, especially older code, e.g. in Linux.
Bonus points if you find yourself actually saying it out loud while typing it.
I have used the word "shenanigans" way more in a couple of years of agentic coding than in 30 years of writing code with humans.
ai llm are doing what i tell them to.
if you’re building something meaningful (in my case a platform used by many people across many companies) you want to ensure you
1. have actual systems engineering and architecture in mind that you want the models to
2. implement based on what you tell it to do
when i was just telling the models what i want done without doing due diligence it would go and do some moronic implementation that was awful. mid input = mid output
these days i just maintain specifications documents and the AI follows everything i tell it to in that document. so when i tell it to dos one thing, the result is made following those architecture specs.
i have code that is single resp, modular, easy to extend and test.
i would ballpark 95% of the time i get what i asked for.
sometimes it tries to be clever in cases that weren’t covered in my arch specs. in those 5% of cases i go and update my specs.
source: used billions of tokens worth to build something actually in production across both mobile platforms and web, deployed on my own cloud infra. i use codex mainly. some claude.
But Claude models seem to be better at long term problems or more ambiguous problems.
I'm curious as to what the primary benefit here. Are there secret improvements in training? There hasn't been much in fundamental model architecture, I don't think. What about harnesses? I wonder what's pushing the AI. It seems like harnesses is the main thing pushing AI ever since CoT.
I think the end game is routed model usage and SLMs. I think Apple is going to prove this in the consumer space pretty handily and I'm curious how the Android ecosystem responds since the hardware is considerably lacking in model performance. I think Apple has a huge opportunity here, as much as I don't like their current ecosystem of walled garden. They did position themselves very well with ARM and custom chips for their hardware. Hopefully the broader ecosystem of ARM and Linux are able to make some headway and we see a more formalized, and broadly accepted, architecture to capitalize on.
I’m sure you could put something similar together with a bunch of duct tape and 2 weeks of effort, but it won’t work nearly as nicely nor out of the box. so…what am i missing?
My company has an agreement with the big providers and while i'm pretty sure they think about how to get budget back, its an competitive advantage and normal people will not learn different model behaviours.
At least for now.
Regardless of what others are doing, US labs here are just rushing to IPO. It's NOT a sign of confidence.
It's the equivalent of saying you have confidence in SpaceX making revenue by renting out their data center (instead of their AI making bank).
On the same note. if spacex is doing datacenters on earth successfully what's wrong with that? They rented cloud infra to a #2 or #3 provider in the world after < 2 years in business. It's a success, no?
If you get hired as a staff engineer and do the work of a junior, what's wrong with that?
Clearly xAI (now part of spaceX) did not raise funds to be a data center. The margins are way different. There are plenty of recent IPOs in that area that are worth at most billions not trillions.
> going to IPO is a sign of confidence , you need to report a lot of things, that private companies don't.
This isn't going to IPO. This is rushing to IPO. It is a sign of confidence that the market or wider environment might crash soon so we need the liquidity now.
> This is an exact reason chinese labs do not rush to go public.
Maybe or maybe not. If you are referring to Chinese labs - both the Hong Kong and China stock market are way weaker than Nasdaq. It's not comparable. Check all the recent Hong Kong IPOs that have tanked.
So no, reason not to might just be: no money in it.
There are huge numbers of users (myself included) that do have an exact idea of what inference costs are - on open models. We can buy tokens from 3rd parties that have no motivation to subsidize our use. That's to say, there's a fair marketplace[1] and we're hanging out there.
If you want to say "I don't think anyone has a firm grasp on actual inference costs on these proprietary/closed models", then I could agree with that.
China subsidizes strategic industries, and they have heavily done so with AI. And DeepSeek specifically has said they have no commercialization plans.
For example: https://www.boc.cn/aboutboc/bi1/202501/t20250123_25254674.ht...
It’s generally established that Anthropic/OpenAI are going for all out performance with big VC dollars at the expense of efficiency and China has geopolitically limited compute and an inventive to compete on value per dollar.
We know roughly how much these companies spend and what their revenues are. Based on that, they'd have to more than double revenue (without spending more money) just to stay even, and that's not good enough given how deep in the hole they are.
> OpenAI and Anthropic are heavily subsidizing their inference -- no wait, they are charging the most they can get away with before going public. Where is the truth?
Both are true. I mean, I'd be willing to spend a bit more than I do now, but not more than double, and neither are most companies. The company I work for is currently investigating how to reduce LLM spend, not looking to spend more.
Both. They are charging the most they can get away with and that amount is still heavily subsidized by VC capital.
Now that 200USD subscription starts to feel cheap...
I haven't gotten close to this either before, but now we wanted to move fast because this branch gets conflicts all the time and we want to get over with the migration asap.
And don't get me wrong. Opus did an absolutely horrible job at first, second and third round in this task. You really needed to steer it to get to the right solution.
And now Fable is out. And its first round of code reviews for this huge PR was definitely worth the money too...
Don't think that I'm just shrugging to that number. I see it every day, and I don't like that it's in the thousands now. But for people paying the 100 or 200 dollar plans, I'm not super sure if you will be able to use them in the future if the token price is in the thousands for a bit bigger task...
If I'd pay this from my own pocket, I'd definitely go with DeepSeek or local models and figure it out how to make the best use of them.
IOW, you don't really think the value of this work is really worth $4k.
> why would I pay to do my job?
The question is: how long do you think that you employer will be willing to pay for you and Anthropic, if you yourself said if it were your money you'd put some time and effort to work with an open model?
I wonder what this question really means? Anthropic is useless if you don't know what to do with it. It's very useful if you do, and you can guide it to do the right things. Yes, it will for sure reduce the amount of people we need to hire. But we are always looking for hires who know what they do and can utilize agents to be faster.
But if you think about how long employer is willing to pay 10-20k per month per seat for Anthropic? I can't see this to be feasible and it will have to end at some point.
It's worth it, and I can afford it, but I am not really the right type of user for token-based usage. It's all for personal and free work.
Unfortunately, that doesn't work within a single session. The K-V cache of a model is intertwined with the model's configuration. Switching models invalidates the cache, meaning everything up to the point of the switchover is processed like a new, uncached input token.
Per Anthropic's pricing doc, an Opus 4.8 cache hit costs 50¢/MTok, while Haiku costs $1/MTok for uncached input.
Model selection works best if sessions are short and self-contained, particularly if the first few interactions can reliably classify the model need. That probably covers most 'support chatbot' use-cases, but it doesn't describe the kinds of heavy agentic automation that really chews through token budgets.
I don't think this is true if you simply quantize the model or run it with fewer active experts? The underlying weights would stay the same. You could also play further tricks with skipping some of the model's middle layers outright, which works surprisingly well due to how skip connections are used.
Most AI companies are just testing the waters with paid tiers right now, their greatest fear with increased pricing is folks reverting back to wikipedia, stack-overflow and other public domain organic activity buzzing back to life; that will kill any RoI potential in LLMs forever. They're playing the wait game instead, observing how the digital sphere reacts to every little increase in price.
If that weren't the case, they'd be pricing at lucrative premiums already and even gotten away in short-term considering the increased dependency in the enterprise world. But that'd be like killing for the golden egg too soon and losing all long-term potential.
Once the folks are so addicted to LLMs that even writing a hello world program sounds like a nightmare and coming up with an article draft feels like reinventing Egyptian glyphs, that's when the real pricing hammer will come.
Anthropic wanting to switch billing to API rates is them just wanting to generate more profit.
Even if subscriptions are locally profitable (i. e., the cost of the subscription covers the cost of inference), they're still subsidized because they don't cover training and running the company; otherwise, these companies would be profitable.
Take a look at China for example - they have no access to NVIDIA, so they're trying to build their own hardware, they have no unlimited funding, so they try to optimize things.
And Anthropic is complete opposite of that - if NVIDIA were to triple their prices tomorrow, Anthropic would still pay them.
In the end, either we all somehow go mad and start paying Anthropic tens of thousands of dollars per month so support this madness, or we will go with whoever isn't lighting cash on fire.
Not true. Stop following US media spam if needed.
1. Very recently, the US did close a loophole on sanctions that allowed Chinese companies to use NVIDIA hardware outside of China i.e. before that was closed they all had access. The trick was train outside, do adjustments, ship the disks back and use non-NVIDIA in China, but at least the training and endpoints not hosted in China could all use NVIDIA.
2. There's been plenty of reports including fines and bans e.g. to Supermicro on smuggling NVIDIA hardware to China. I doubt it has been stopped. You can't catch everyone.
Granted, it could still mean that Anthropic just chooses to lose money - but that's Anthropic's choice.
DeepSeek has proven that inference can be much, much cheaper than what Anthropic advertises on their API rates page.
So they are profitable?
I think you are mismatching accounting terms.
You can't say the 'subscriptions' are profitable without accounting for the cost of making the model that is the source of the subscription.
They are heavily subsidized by the shareholders. Investing, running at a loss, with hope of some future profitability.
If saner factory can sell you the same tool at a fraction of the cost of a gold plated factory, your choice is going to be obvious.
Having said that, I found the cloud dev environments slow to the point where I wasn’t sure if it had frozen, so I never looked back.
Though the day is coming when there’s no distinguishing, I’m sure.
Also, is it really a defense department when you're starting wars of aggression every 15 years or so?
Just like how changing Kennedy Center letterhead to Trump Kennedy Center for a year didn't actually legally rename it.
Once a case with sufficient standing got in front of a judge it reverted to the actual legal name on the basis that only Congress can change the statutorily defined name.
For an admin so obsessed with legal names instead of chosen ones, you’d think they’d be less hypocritical.
I'm doing basic web development here utilizing animejs. Nothing too complicated (mostly saving time doing the scaffolding, still write the bulk of animations manually).
Truly believe that American companies are going to get completely curb stomped by China due to greed, ineptitude, and violating the social contract.
Deepseek V4 Flash is suprisingly capable and insanely cheap. It takes so much to get the session cost to get to $0.01.
I agree with you on pricing, but what do you mean by this?
Why aren't corporations doing more to help workers with childcare? Why aren't they doing more profit sharing with workers? Why aren't they encouraging unions or sectorial bargaining? Why isn't the government mandating any of this?
Americans very rarely benefit when US corporations do well. That needs to change. No one benefits if Meta continues making billions in profit every quarter while society suffers from isolation, depression, suicide, and scams from their services. Americans don't benefit if health insurance companies are making massive profits while they can't afford deductibles.
Our society has been setup to simply extract wealth in all facets of life. That's a sick society and it needs to change.
I'm not saying China does this better, in fact China has some of the worse worker rights out of all the industrialized countries; but at least American consumers would benefit from cheaper higher quality Chinese goods. The world would likely benefit too if America got off the cold war hype train that did nothing to benefit humanity outside of those making weapon systems.
The AI companies sure are a brilliant example of corporations needing to do more to help their employees pay for childcare.
I am on the $100 Max plan.
I do wonder if you switched models mid-session, you would have lost all your cache. Reloading the context into cache can really eat through your usage.
I had it analyze a project I was working on with Opus 4.8, and it blew through 23% of my session limit in one go. Does not portend well for my budget.
> Fable 5's safety measures flagged this message for cybersecurity or biology topics.
> They may flag safe, normal content as well.
> These measures let us bring you Mythos-level capability in other areas sooner, and we're working to refine them.
Here are the results of the agentic code review session:
┌──────────────────────────┬───────────────┬────────────────┐
│ Agent │ Fable 5 turns │ Opus 4.8 turns │
├──────────────────────────┼───────────────┼────────────────┤
│ values │ 134 │ 0 │
├──────────────────────────┼───────────────┼────────────────┤
│ data-intrinsics │ 104 │ 0 │
├──────────────────────────┼───────────────┼────────────────┤
│ tools-tests-build │ 81 │ 0 │
├──────────────────────────┼───────────────┼────────────────┤
│ core-intrinsics (failed) │ 25 │ 0 │
├──────────────────────────┼───────────────┼────────────────┤
│ system-memory │ 44 │ 20 │
├──────────────────────────┼───────────────┼────────────────┤
│ reader-modules │ 104 │ 25 │
├──────────────────────────┼───────────────┼────────────────┤
│ linux-startup │ 95 │ 15 │
└──────────────────────────┴───────────────┴────────────────┘
They also, FWIW, say that they've instituted new policies on their end such as logging any human access to the stored data and automated deletion after 30 days in "most" cases (with another link to a document detailing that further).
Assuming this isn't just a supply issue on their side, nothing says "ethical AI" like only allowing mega corporations to use it through cost barriers.
How many government sanctioned school bombings does it take for them to quit working with said government? For now we know that number is somewhere between infinity and 1.
The question of collaboration with USG is a much more complex one, but is not the one raised above.
Edit: I'll also add that I doubt any AI-doom people "trust" Anthropic per se. The entire angle of questioning – again – misunderstands the AI-doom argument. You appear to think that if companies behave unethically, they cannot be trusted and they will not produce good outcomes, inversely: if they behave ethically, they can be trusted, and they will produce good outcomes.
Any competent AI-doomer would argue that ethics or trust are essentially irrelevant.
The entire problem is that people can act totally reasonably, even ethically, and this is not a guarantee of good outcomes. Situations can be created in which completely ethical, reasonable behavior actually produces a bad outcome. You do not need to assume people are bad in order to produce a bad outcome, and inversely you cannot assume that you will get a good outcome from good people.
"Arms races" are one class of situations that often have this characteristic. "Bureaucracy" is another class that we encounter a lot in daily life. There's a lot of them!
Anthropic needs to be at least somewhat in the good graces of a capricious administration that is already under pressure from businesses and citizens to regulate AI companies across multiple different domains, whether it's energy consumption, job displacement, military and defense applications, surveillance, etc.
If Anthropic wants to survive, they need to acquire influence with the government that most impacts them as an American company, and a massive exporter of services in the AI space to other countries, otherwise they could get locked down and locked out of the market for national security reasons.
It sucks, but sometimes the survival choice is to make an ethical compromise in hopes that you can still be around to make better decisions later.
This "simple" fact needs quite a bit of additional context and work. Making grandiose ethical claims like this can be countered with other grandiose claims such as the fact that there is no ethical existence under communism or socialism.
The fact that there is no ethical consumption under capitalism is not material to whether or not ethical existence is possible under communism or socialism. In order to survive in a capitalist society, one inherently has to make choices that require trade-offs, and those trade-offs are burdened by a history of decisions made not just by the people alive today, but our ancestors as well. Does that mean I walk around chanting "Reparations", "Land-back", or other calls to action? No, but I do acknowledge that there are unresolved issues and as a Canadian, I know we need to do more to resolve treaty issues, and environmental issues, and system discrimination. I also know that Americans need to do better to address systemic discrimination and many, many other issues. It also doesn't mean I want to give back my house, or give away all of my possessions. It just means I try to make good choices and support businesses and people that are open about the trade-offs they make and try to engage as ethically as possible.
Acknowledging those facts doesn't absolve us of responsibility, it's a framework that allows folks concerned about whether or not they are doing the right thing to accept the trade-offs that they choose to make and be responsible and accountable for those choices to themselves or their communities.
We live in a world with scarce resources. It's possible that with a foundational redesign of the global economy, and the requisite authoritarian government that would be required to force such a redesign, we could eliminate food scarcity, solve energy scarcity, and make sure that everyone has a place to live. Those trade-offs are probably not worth the ethical cost in political and physical violence required to accomplish it. We have seen the trade-offs that happen when the powerful are able to exploit communist or socialist governments. We are seeing the "late stage capitalism" impacts of allowing the powerful to exploit capitalism in democratic societies. Acknowledging that the current capitalist system has lead to the greatest prosperity for the upper echelon (financially) of humanity, and a dramatic reduction in global poverty shouldn't obscure the reality that much of that wealth comes from exploitation of people and the environment.
It's a huge problem to unwind, and we can't let the burden of every choice that we make stop us from trying to do better, but we (as in society in general) can't do better if we don't at least acknowledge the compromises we are making along the way, and try to plan to fix it in the future.
Probably a topic better suited to beer and a pub setting than HN though :P
I don't believe that this is a fact. How are you demonstrating that this is a fact?
When you talk about things like reparations or "land back" you're already cargo-culting in concepts and ideas that themselves need to be fleshed out in order to make a subsequent claim that a specific economic system is unethical. Someone can just argue all economic systems are unethical, how are you going to defend against that? And can you pay reparations for example without going back in all of human history and finding all cases of injustices and then tallying it up? Why pick an arbitrary point in time? Better yet, why not start in countries where slavery still exists instead of focusing on the west which led the world in abolishing slavery and created concepts such as universal human rights.
Even with respect to "eliminating food scarcity" - eliminate in what sense? All olive groves and grapevines and rice farms have to be destroyed and rebuilt to only build certain foods?
Dabbling in communism or other inhumane and authoritarian governmental systems is extremely dangerous and in the same vein of extraordinary claims required extraordinary evidence, suggesting as you did creating an authoritarian government to create a utopia is precisely the same project of suffering and death that mass murderers throughout history have undertaken to abject failure, and thus, you need some incredible amount of evidence and theory to be able to even fairly suggest going down this path.
I am not going to do the work of gathering the evidence for you, and I don't think this is the right venue for a debate on the topic.
If you don't have evidence I think it's mature of you to admit that and applaud you in doing so. We all like to just talk and don't have to always provide evidence for every citation or what not and it's fair to just say hey I'm just making this up and it requires further discussion.
Talk about a strawman!
(I’m highly confident open models will eventually achieve a similar performance benchmark with distillation over time)
AI Savings Misses 'Should Be Making Executives Uncomfortable,' Bain Says - https://news.ycombinator.com/item?id=48359010 - June 2026 (0 comments)
AI sticker shock hits corporate America- https://news.ycombinator.com/item?id=48307098 - May 2026 (146 comments)
ZIRP (zero interest rate policy) is over, software engineers no longer call the shots now that there isn’t vast amounts of capital chasing yield, and that capital bidding up salaries and keeping the labor market for engineers tight.
If you are x more productive with generative AI, very shortly you are going to have to prove it with a token budget (or, if you’re lucky, an org willing to spend for on prem hardware for capped token cost, fixed capex vs uncapped opex).
The comparison is not SWE vs SWE with AI. It is SWE vs SWE with AI with a constrained token budget ($x/month) delivering the same value at the same or lower cost. If you cannot prove that you are wildly (vs marginally) more productive with the AI, why would they pay for it? Prove it.
https://abhishek-shankar.com/posts/ai-coding-bill-headcount-...
> That is the real content of the Uber story, and it is why filing it under "budgeting discipline" misses what is actually unfolding across half the engineering organizations in the country right now. They ran the same experiment Uber ran, most of them without Uber's $3.4 billion R&D cushion to absorb the surprise, and almost none of them having modeled the heavy-user tail or instrumented the gap between tokens consumed and value shipped. The reckoning will arrive for each of them on their own fiscal calendar, and the first instinct will be the wrong one. The tool is too good to abandon, the bill is too large to absorb, and the only durable resolution runs through a question the entire rollout was designed to defer.
> You cannot get labor-replacement economics out of a tool you deployed as a labor supplement, and the bill comes due before anyone is willing to admit which one they actually bought.
I just use dumb and fast models now. I'm more engaged. I think that the higher the quality of the model, the more you tend to vibe with it, and then the more hallucinations you then miss. I'm not sure which is more productive, but I definitely burn out faster the more I vibe. At some point you're spending your time on forums, discord, or youtube instead of engaged with what you're building. Or you yak shave about your tooling and end up creating the 600th multi-agent gastown harness and blowing thousands of dollars on tokens to create it only to discover it's too expense to actually use.
Upd: I meant big picture, not with respect to this model release. Where do subscriptions figure into their strategic vision. Will consumers end up paying enterprise prices in the future?
Why wouldn't Anthropic just wait until people start subscribing, do some kind of marketing push, or obtain some kind of other sustainable revenue stream, before they go IPO? I wonder if they see the writing on the wall with all of this and want to cash out as quickly as possible?
Specifically they need businesses that fired people and adapted their business to the products, so when the unsubsidized costs hit the businesses are forced to eat the true costs.
Yes they can't afford to give the products for free, but what is essentially happening with AI services is economic dumping, keep costs artificially low to get people to fire everybody, and then Jack the rates once they have Monopoly control
I agree. They need addicts, but they are high on their own supply and everyone else can see the danger in getting hooked.
They'll probably tighten the quotas to reign in whales though.
Realistically I think Anthropic just has insane demand but finite capacity to run models, and Fable will just make them more money if they dedicate it to API pricing. I suspect the goal here is something like: get individual engineers/PMs on their personal plans to taste Fable and then go to their meetings and say "Yes doubling the price of every single input/output token is a good idea, boss".
The only reason why I pay $200 is because LLM's errors costs me that much, at worst. If "make no error" starts working - sure. But surely, unless you have millions of dollars of cash to burn, a coin flip that costs $5000 is an insane idea?
Going PAYG only will effectively take these tools away from a huge amount of people and accelerate the push for local LLMs.
OTOH, accelerating the push for local LLMs would also be fine with me.
The AI landscape is changing rapidly, and with Apple announcing the option to change the AI backend, and potential requirements enable AI choices as well, similar to EU browser choice requirements (this is more reading tea leaves than any actual requirements I am aware of). The new OS changes coming to support Googlebook, and deep Copilot/AI integration into Windows will make maintaining user facing subscriptions essential for independent model developers like OpenAI, Anthropic, and Mistal to remain relevant longer term.
If the don't maintain that relevance there is increasing likelihood that they will get consumed by other companies whether it's Apple, Microsoft or Google to form a foundation for their OS, or other cloud providers.
It's kind of annoying not getting access to the primo model and paying 200 bucks a month. I understand 200 bucks a month is basically nothing though.
Like I don't totally understand why they'd let me have it for a couple weeks and then take it away and say I can have it but I have to pay retail and retail is like $1,000 a day.
It's better to have loved and lost than to have never loved at all??
As a consumer I can choose to buy subscriptions to a range of things, including $5 droplets or VMs on a broad range of cloud hosting providers. I can even buy cheap bare metal at a bunch of providers at an affordable retail rate.
I can also buy "unlimited" AI packages that will be optimized to fit the cost model from a variety of services, with different impacts, such as rolling outages when I consume a daily or hourly allotment.
Right now VC and the investor class are subsidizing the rapid evolution of the services and availability, but that VC is running out. In more traditional economies, AI would have developed and rolled out more slowly, and through metered subscriptions, with the eventual rolling out of "unlimited" packages like telephone, internet, or cell services once the market became commoditized.
We have seen a big inversion of that with the race to "win" AI marketshare. Now the true cost is being exposed, and the most competitive and capable models are hideously expensive to operate, so it makes sense that we are moving to metered billing for a utility service. If you want gas, you can buy regular or premium. If you have a premium car you definitely want the premium, but for most people regular is good.
Give it a couple of years, and the survivors will settle around fairly industry standard models of consumer grade services, pro-sumer accounts, and business/enterprise models.
Things are still shaking out, but I get the sadness. Luckily I work at a big tech company who is banging the drum on doing experimentation so I use my prosumer claude pro and other accounts at home for hobby stuff, and save my heavy lifting and potentially experimentation for work :P
Opus 4.8 produces output in 15 minutes that is 3-4 hours of my work away from output that used to take me 40ish hours (a solid week of dedicated effort).
Last year(-ish, maybe it was 18 months, I forget when the jump happened), the frontier models couldn't touch this work. The output looked like a hardworking intern on their first day. Nice formatting, decent volume of words, but no understanding.
So it might work if it turns out to be a substantial leap in capability.
The newer models are smarter but really ficklle and hard to get meaningful work out of
4.6 was a workhorse
Probably all about the IPO.
It's the same exact speed as opus >=4.5, sonnet 4.5, and twice the speed of opus <=4.1
It must have about the same active parameters, or else its a larger model running in turbo mode (smaller batches) and being heavily subsidized for some reason. But given most of the benchmarks are within 5% I doubt it is a much larger model. Most perplexing.
I think they might be hitting a point where subsidizing the expensive models for subscriptions makes less and less sense.
With Opus 4.X, last month I paid 100 USD for the Max subscription and got a token equivalent of 4.1k USD.
I imagine that Fable is more expensive to run.
As annoyed as I am about this move, I get it. Users flood the newest, best model whether they really need it or not, and are efficient at using their entire quota. They've had so much trouble reigning in subscription usage it makes sense.
The step-up in intelligence looks massive (we'll see in practice), but the price is getting to a point where it's making me question if it's even worth giving it a try.
Good competitors will probably be out soon, which should level the playing field. I am more excited about that, just the fact that they showed that such an improvement is possible. I'm okay waiting a bit longer for this to become attainable for plebs like me.
Kind of like billing a programmer by the hour.
Perhaps not that close to US salaries, but those are inflated to hell. Worldwide senior engineers and scientists have salaries just about an order of magnitude away from AI subscriptions that you can use most of the day every day.
Do we know this? I’ve seen evidence they lose money on heavy users. But so do gyms.
Most gyms sell more subscriptions than they can fit under their roof at one time. If a gym only sells to heavy users, it will either be constantly turning members away or have to buy more equipment. Its equipment will wear off faster. Depending on amenities, it will go through towels, soap, water, et cetera faster, too.
Unless they're really, seriously wasteful with the soap.. there's no chance a gym is losing money on a heavy user
Right now all these AI subscriptions are priced like Planet Fitness, but they're used like Equinox. They're hoping that the new a la carte offerings will move their pricing more in that direction as well.
Where?
What I wonder however is if these tools will become something I use at work only. $100/month is already a massive stretch budget wise. If these models keep devouring tokens there’s no way I’d get the same usage time out of them for $100 in usage credits.
I just don’t think I’d use them much at all at home.
...
Unlike our interventions for cybersecurity, biology and chemistry, and distillation attempts, these safeguards will not be visible to the user."
[edit] -- I see that this comes from the system card -- dang merged the comments from the other discussion so that explains the confusion.
If you rely on this as a core part of your business/profession, you will be at their mercy and subject to whatever whims or challenges they have.
> Fable 5 · Most capable for your hardest and longest-running tasks · Uses your limits ~2× faster than Opus
Pay-as-you go isn't a common thing in SaaS. For example, except for AWS SES, all email providers are bulk-subscription based.
Sounds like "bait and wait".
If you think about it, the more people pay for these new and more resource hungry models, the longer it takes for them to become no extra cost and the longer it takes the more people are tempted to pay extra.
If you have good expertise in a domain and access to cheaper models, you may still be more skilled than someone without expertise but a lot of money to bruteforce the problems using SOTA LLMs.
Of course, they are a casino as well giving you free spins at the wheel with their new Fable machine, and it is done on purpose.
Once there freebies have expired, many of its users will begin to gamble more on the new casino machine and will realize that it is expensive.
The ramifications go beyond the individual which is why I assume they mentioned it. They don’t need to use it/not use it for it to have interesting implications.
Is it nice we get the trial? Sure. Is it also a common play in the playbook of tech companies? Yes.
Anthropic does not care about us and isn't going to talk to you either and will extract from you as much as possible.
The true answer is local models.