> Siri Recap runs only when you choose. You can turn on Siri Recap in the Watch app on iPhone. When you turn it on, you can set a schedule based on time and/or location (for example, "only at work" or "never at night"), or you can manually turn it on and off. Across all these options, you can always turn it on or off at any time from Control Center.
> Siri Recap and Live Rewind help jog your memory, not provide a detailed record. Siri Recap doesn't produce a transcript. It generates a brief, high-level summary — comparable to notes you might jot down after a conversation. Live Rewind shows a text snippet of what was just said in the last 15 seconds, so you can catch something you missed.
The only reason OpenAI and Anthropic are making a loss is due to training new models to keep up with competition - not because the industry is flawed in terms of business model.
> The only reason OpenAI and Anthropic are making a loss is due to training new models to keep up with competition - not because the industry is flawed in terms of business model.
So in other words they are not profitable? Like you cant say they are profitable and then in the next sentence say they make a loss. That is not how profit works.
The point is that the model is already immensely profitable. There is no fundamental reason why AI isn't profitable. It already is.
Insane competition does not last forever. When chip manufacturing first started, there were dozens of companies that had fabs that made compute chips. Nowadays, only TSMC is viable. Samsung and Intel survived because of geopolitics.
> Is that not self evident by the insane revenue from frontier labs?
No...? Of course not?
Because revenue is only one side of the equation. Did you ever look at total cumulative OPEX and CAPEX, and how long it will take them to even just break even at current growth?
They're likely over $80b ARR by now. They'll be at $800b ARR next year at the same rate. Let's say their growth gets cut down to 3x instead of 10x - that's still $240b ARR by this time next year.
When you are growing so fast, you don't need to make a net profit. You just need to make sure your unit economics are good - which it seems like they are given reports that their gross margins are at 60-70%.
> They're likely over $80b ARR by now. They'll be at $800b ARR next year at the same rate.
And they will be 800 trilion ARR in a couple of years, following that same rate! 8 quadrillion by 2029!
> When you are growing so fast, you don't need to make a net profit. You just need to make sure your unit economics are good - which it seems like they are given reports that their gross margins are at 60-70%.
If their margins were anywhere near this good, they wouldn't need to raise so much money so often.
If you create a machine that turns 1 dollar into 3 dollars, you don't dillute your ownership of the machine, you use your fabulous profits to expand your machine's capabilities.
If their margins were anywhere near this good, they wouldn't need to raise so much money so often.
Why not? They are reinvesting into growth. There isn't a clear winner yet and Anthropic wants to make sure it is one of them. Taking a profit now while letting OpenAI take your marketshare and train better models is not very smart.
Or they bleed money like crazy, and their margins are pretty awful. Which is the correct answer.
Your $200 subscription is a major net loss for them. The vast majority that pays for that would cancel in a heartbeat the moment they had to pay API prices. Which may or may not be profitable, I am not entirely sure. But for the sake of argument, let's assume that it is.
Wihout insight on how much enterprise is paying, it is impossible to draw any conclusions. Unless you have any access to their contracts and are willing to share evidence? I find that highly unlikely.
People here throw around crazy numbers - the dude above was claiming they have some insane good margins, numberd that he took out of his ass.
The only evidence I have is that they are incredibly unprofitable, and they keep raising insane amounts of capital like crazy.
There was a leak sometime ago that they were EBITDA positive during a quarter where they didn't pay for part of their compute. And EBITDA is a cute metric to use when depreciation is actually very important to them, as a model from a year or so ago is nearly worthless.
serving models is very profitable (70%+) but the issue is you need to invest in training the next iteration. but so far all of anthropics models have been profitable fully loaded
the vast majority of the labs revenue is from enterprise api usage (theres public sources from the information and ramp). but the risk there is customer concentration, where most of the revenue comes from other tech companies and a chunk of it is from foreign labs distilling
so i am drawing a conclusion that the labs' business model is good, maybe not as great as boosters think it is. if they make real progress on the biosciences like drug discovery that could turn it into an amazing business
This is not evidence. This is random people speculating on Anthropic's margins without any real evidence.
Just because it is on some blog post, it does not make it true.
I wasted the time to read the first blog post. It considers 100% utilization over the course of years to calculate an estimation, and it did not consider depreciation for the model itself. That thing is extremely extensive to create, and after a relatively short amount of time is considered outdated.
Are we still calculated $200 subscription token spend based on their highly inflated API token cost and then concluding that they must be losing money on all $200 subscriptions?
Sure, there is revenue, and market valuation. Is there _profits_ in frontier models ?
What is the horizon for openai and anthropic to _make_ money ? Will they achieve that by charging more for frontier models, or investing slightly less in training frontier models, etc... ?
> we know better for you so we’ll adjust the behavior accordingly
9/10 times I've used this setting, what I've actually wanted was to temporarily change the behavior. This just saves me from the need to remember to undo it (which I won't remember in a timely manner).
This isn't "we know better than you" paternalistic control - the default/easy behavior should cater to the actions and situation we do regularly. I'm sure they do, in fact, know that for most of their customers, when they disable wifi it's because that user wants to temporarily disconnect from the current wifi, but will want to use wifi shortly after, but won't necessarily remember to go into settings and re-enable.
It should be symmetrically facile to do both behaviors, assuming there's not an invisible thumb on the scale that wants to procedurally force you to make the choice most in the company's interest. But its not.
It is more effortful, more complex, and requires greater care to actually get Wifi and Bluetooth to the OFF state than ON, or ON+CONNECTED. Respectfully, that is backwards, it should be the easiest thing to turn something off.
Yet, here there is needless nuance added because there is a benefit to the company that the device is either chronically or as much as possible connected to the internet, connected to bluetooth, low battery mode is off, and if they cant have it connected to the internet and bluetooth, well, at least keep the beaconing on and sending out signals and at least meshing so advertisers and stores and their own Find/lost services are still able to benefit.
Thats not even getting into passive Fingerprinting and secondary furtive location tracking. I was surprised to learn that having bluetooth on identifies your device to stores and anyone else who can track bluetooth devices in range. Even when you toggled off bluetooth, you're still being profiled. I wonder if actually going into settings and turning off bluetooth or wifi even totally stops the device from continuing to signal in all ways, it would be unsurprising if thats another UI nicety that doesnt actually mean what it says or implies
> It is more effortful, more complex, and requires greater care to actually get Wifi and Bluetooth to the OFF state than ON, or ON+CONNECTED. Respectfully, that is backwards, it should be the easiest thing to turn something off.
Why?
I get you have some principled opinions or deep belief about tracking or betrayal from corporations or something. That’s nice.
But why should it be easier to do the things that people don’t want to do often instead of the things that people do want to do often?
Good UI makes repeated actions easier than rare actions. That is good UI because it makes the common action easier, conspiracy theory about tracking and fingerprinting be damned.
The cooler thing is that if a substantial amount of the 216M TV owners tried to arbitrate across the country/world, they'd cave so fast. Arbitration is cheaper than an individual law suit, but much much more expensive than a class-action.
Because violating the law was the prerequisite to starting their business, without it they're worth $0 and have no models. They've already survived Training on customer input may help the models but it isn't "bet the farm" helpful. Now, they have a thriving business, so they shouldn't risk their business for incremental data that they can buy.
At this point, the reputation of the business matters too. Fable explicitly didn't support zero-retention usage, and it saw significantly lower adoption vs other flagship models, and their past releases. Being caught abusing enterprise contracts is really hard to dig out of.
and they're under no contract with those millions of websites and books, but it's a whole other thing when it's a paying customer, especially a large enterprise with a signed contract with DPAs etc.
That's the thing though. There are very few "actual workflows" in many people's lives (especially if you exclude their job), and many many "one off (few off) workflow". That's the magic.
So many tasks I need to do once, or just a few times ever, but they build up. While I'm not sure I'd use it for immigration forms specifically, that's the kind of task that's tedious and done rarely, so a dedicated tool doesn't help because who would be familiar with that tool and have it handy?
I’ve been super pro-Luna lately. I really hope that Gemini-Flash-Lite is positioned to compete with it. We all know that Anthropic has abandoned Haiku and it would never be that cheap.
Probably shouldn’t say this here but I’ve been planning to up my $20/mo exploratory ChatGPT subscription to the $100/mo tier as soon as I hit my cap. Between the progress and quality of Luna and their continuous resets, it’s been a few months now that I’ve lived off the $20 tier, frankly waiting for the need to upgrade, credit card in hand.
I’m always trying new models, like many of us here, but the price is just so good for a well balanced, American, hosted model.
How much of that 16mo is design versus just production? If there was a “plug and play” chip where you just BYO weights, how long would it take?
The bigger issue seems to be that these chips can’t hold that many weights at the moment.
(I’m curious if chips with large weights in them would be more tolerant or less to yield issues. If you flip a few bits in the weights, does it really matter at scale?)
Talaas, from what I understand is building stuff just like that. The infra is the same and the weights layer is all you need to change. I guess you could half etch the chips and then finish them with the weights only. I think their turnaround is 6-8 Weeks. The size of the models fitting on the chips at the moment is llama 3 I think?
Those jobs aren’t coming back as-is, because AI isn’t going anywhere. Job markets change over time.
I’ve just started transitioning my career closer to management, so I develop/manage roughly 50/50. At this point, if I had to manage a tight budget, I’d rather have an entry level salary as tokens for a senior engineer.
We’ll need a way, as a society, to help new adults transitions from education to productive employee. But individual companies don’t have aligned incentives.
> I’ve just started transitioning my career closer to management, so I develop/manage roughly 50/50. At this point, if I had to manage a tight budget, I’d rather have an entry level salary as tokens for a senior engineer.
A high-potential junior you can at least justify as an investment in the future.
The other people who are going to get squeezed are the later-career developers who are below-average in ability. (I'm not using "below average" here pejoratively - just noting that by definition, half of developers are below-average.) Outside of big tech, there are piles of organizations with devs who don't make big-tech salary and who are only marginally productive for what they're getting paid.
If you're an organization that only has the budget to pay non-junior devs $80k... you can see the implications of having the option to spend your dev budget on seniors + tokens instead.
I'm not seeing the reason why typical engineers can't vibecode. Non-coders can vibecode. I have read hundreds of posts claiming that senior engineers can vibecode but not a single one saying that juniors cannot.
I'm not saying they can't, but rather that with the time seniors then spend cleaning up their tricky bugs, architectural problems, etc. the seniors could have just done the work themselves.
It won’t work. I have nothing against taxation and funding education but the issue isn’t that type of education.
If you work in the industry in a well funded company (the ones who’d be paying these taxes), then you probably have more college-grad applicants than can hire. Probably more than you can hire from “good” schools, some with masters degrees, etc. Maybe education should be subsidized, but theres no lack of volume of people with the formal education today. Top schools (Berkeley, MIT, etc) today are reporting growing unemployment rates among their CS grads. We have more than enough students.
But that formal education isn’t the necessary skill missing. It’s understanding how to structure your work. Learning how to understand the complexity of a problem and estimate how long it’ll take. Learning how to spot future issues during code review (not just bugs and not just optimizations, but “one way doors”, future migration headaches etc). Learning architectural patterns and common prior-art. Even learning to show up on time, be polite in meetings, and navigate office politics.
There are just skills you need to see and practice to understand - you can’t teach them in a classroom. That’s why electricians have apprenticeship and doctors have residency, lawyers clerk and programmers had internships. The difference is that before AI, a young grad with an internship learned enough to answer basic bugs and learn on the job. But now that AI can fix bugs and small tasks faster than a person, the minimum skill threshold has exceeded the on-job training standards.
As a software engineer, I selfishly hope that they spend more effort on non software tasks since I’ve feel like we hit a sweet spot where engineers still have some value and autonomy, but a super charged tool.
Pragmatically, I suspect that “non software” tasks will be a tarpit because most tasks can’t be automated and verified as easily in an RL loop compared to software projects. Especially since most skilled labor is either not nearly as expensive as software engineers (eg biologists), or regulated (eg doctors, lawyers).
I suspect the focus will probably shift once software engineering is no longer the biggest cost center for most AI company's clients, and we'll start working on getting rid of the next cost center.
“Siri Recap”
https://support.apple.com/en-us/148354
reply