I think it’s pretty clear what the purpose of this stuff is: get people so invested into the Claude ecosystem with certs and “modernization kits”, so that when the subsidies end and subscription costs shoot up they feel they’re in too deep now to switch to something cheaper.
> so that when the subsidies end and subscription costs shoot up
Subscription costs are capped to API rates as their ceiling (and, realistically, way lower than that - why would you even subscribe if you could just go pay-what-you-use instead), and those are already at a big margin for Anthropic. What still costs them a fuckton of money comparatively is training, but that is only going to get more efficient with more purpose-built hardware on the way.
Basicallly, I don’t see much of a reason to hike subscription prices dramatically. I don’t think they’ll stay at $100/$200 but anyone who’s paying that already knows how much value they’re getting out of that and probably wouldn’t mind paying more.
I'm not sure what you mean, if you max out your subscription perhaps? If you pay $100 and don't use it, you don't get refunded $100 because it's 'capped to API rates' which would've been 0.
He means that anthropic cannot increase the price of the sub because the users can just switch to the regular API pricing which consequently puts a ceiling on the cost of the sub.
Nobody would use a $1k sub if using the API pricing would only cost $500 for comparative service.
For the record, I'm only explaining what he put forward.
I don't agree with the opinion, mainly for two reasons:
The API cost can be increased in conjunction, hence the ceiling is just as variable
The harness is even more important then the model ime, and Claude Code is getting better every month. Even though the alternatives are getting better too, they're at least currently significantly worse IME - I'd say at least 3-6 months behind (compounded by the model, ofc).
And as a third point, unrelated to the original argument: there is no way anthropic is actually treating the sub as a loss leader. It is not cheap. It's only cheap compared to their API pricing, which they can freely set however they want. Compare their pricing to free models like Kimi k2.5 etc. I sincerely doubt anthropics model costs more to run then theirs, and they're profitable at 30% of the price anthropic charges.
> He means that anthropic cannot increase the price of the sub because the users can just switch to the regular API pricing
Not that they cannot increase the price, just that there's a cap on how high they realistically can go. Sure, they can always hike API prices to compensate, but I think people are seriously sleeping on open models these days, because…
> *The harness is even more important then the model ime*, and Claude Code is getting better every month.
…I fully agree with this, and that’s actually the other reason why I don’t think we’ll approach predatory pricing. Right now, the moat is still mostly the model, but as open models improve and become more capable, this is quickly going to shift.
And the truth is that Claude Code just isn’t that great of a harness. Anyone who uses an open-source harness and optimizes it for their personal, individual workflow will quickly realize this. And I’m not even blaming Anthropic or the CC team or calling them incompetent; they are in the unenviable position to have been trailblazers. There weren’t any comparable tools before CC that they could’ve learned from.
The future lies in harnesses that are multi-model, extensible, and have full access to and control over the model’s API, context, and system prompt. Claude Code has none of those things. You can only ever bend it into a shape that approximates your workflow; you can never use it as a tool that natively supports it.
Oh, on that we can agree on! I was using opencode for the last few months, the main reason I went back to cc was for opus, and me preferring the sub over regular API pricing as I'm not using it professionally, only as a hobby. (At work I'm constrained to Copilot. Which is fine at this point, not great but definitely improving - esp. when run as CLI)
I am still hoping for a local first model approach with voice command to generate the main prompt which starts of the plan mode.
Like interactively going through the project while pointing at files or in the UI and possibly browser via the mouse and explaining while "talking" with a dumber but super quick model that acts as a questioner, to wrap things up with higher latency over the wire with the highly capable models.
I suspect that approach is still a few months to years away from viability for latency reasons, but I'm definitely looking forward to that UX
Now huge amount of investment pays for training. This investment expects some returns, to be able to both turn profit and continue the training, rates must be much, much higher.
That is the biggest threat - and likely where things will end up eventually… it’s when that “eventually” is and what the server based providers can pivot to in that time.
This will probably happen unless the industry conspires to roll back the availability of general computation so common people can only own computers with enough power to be glorified thin clients. The way this might look is good hardware never officially being banned, just priced too high for anybody to afford, and produced in small quantities to keep it that way while all production shifts to making massively expensive powerful hardware for corporate buyers.
Seems unlikely. We're already seeing specialized hardware optimized for LLM performance (taalas, groq, cerebras), and simple economies of scale result in these sorts of products being a better value when rented from a server vs purchased/managed/upgraded for the typical the user.
Frontier models will continue to be either exclusively available from servers or significantly more affordable from servers vs local alternatives for the foreseeable future.
> I think it’s pretty clear what the purpose of this stuff is: get people so invested into the Claude ecosystem with certs and “modernization kits”, so that when the subsidies end and subscription costs shoot up they feel they’re in too deep now to switch to something cheaper.
Do you have a source for that? Certainly things like compute and other services that I'm aware of are objectively cheaper, so I'm curious what has gone up.
it’s crazy how you could easily lie about having 10 years experience because your results are not that much different from someone who has only used Claude Code for like a week.
I think the older AI users are even held back because they might be doing things that are not neccessary any more, like explaining basic things, like please don't bring in random dependencies but prefer the ones that are there already, or the classic think really strongly and make a plan, or try to use a prestigious register of the language in attempts to make it think harder.
Nowadays I just paste a test, build, or linter error message into the chat and the clanker knows immediately what to do, where it originated, and looks into causes. Often times I come back to the chat and see a working explanation together with a fix.
Before I had to actually explain why I want it to change some implementation in some direction, otherwise it would refuse no I won't do that because abc. Nowadays I can just pass the raw instruction "please move this into its own function", etc, and it follows.
So yeah, a lot of these skills become outdated very quickly, the technology is changing so fast, and one needs to constantly revisit if what one had to do a couple of months earlier is still required, whether there's still the limits of the technology precisely there or further out.
I’ve no idea what you’re using, but “I simply paste” isn’t giving me good results at all.
An hour ago Gemini decided it needed to scan my entire home folder to find the test file I asked it to look into. Sonnet will definitely try to install new dependencies, even though I’m doing SDD and have a clear AGENTS.md.
I’m always baffled at people’s magic results with LLMs. I’m impressed by the new tools, but lots of comments here would suggest my Gemini/Sonnet/Opus are much worse than yours.
> I think the older AI users are even held back because they might be doing things that are not neccessary any more
As the same age as Linus Torvalds, I'd say that it can be the opposite.
We are so used to "leaky abstractions", that we have just accepted this as another imperfect new tech stack.
Unlike less experienced developers, we know that you have to learn a bit about the underlying layers to use the high level abstraction layer effectively.
What is going on under the hood? What was the sequence of events which caused my inputs to give these outputs / error messages?
Once you learn enough of how the underlying layers work, you'll get far fewer errors because you'll subconciously avoid them. Meanwhile, people with a "I only work at the high-level"-mindset keeps trying to feed the high-level layer different inputs more or less at random.
For LLMs, it's certainly a challenge.
The basic low level LLM architecture is very simple. You can write a naive LLM core inference engine in a few hundred lines of code.
But that is like writing a logic gate simulator and feeding it a huge CPU gate list + many GBs of kernel+rootfs disk images. It doesn't tell you how the thing actually behaves.
So you move up the layers. Often you can't get hard data on how they really work. Instead you rely on empirical and anecdotal data.
But you still form a mental image of what the rough layers are, and what you can expect in their behavior given different inputs.
For LLMs, a critical piece is the context window. It has to be understood and managed to get good results. Make sure it's fed with the right amount of the right data, and you get much better results.
> Nowadays I just paste a test, build, or linter error message into the chat and the clanker knows immediately what to do
That's exactly the right thing to do given the right circumstances.
But if you're doing a big refactoring across a huge code base, you won't get the same good results. You'll need to understand the context window and how your tools/framework feeds it with data for your subagents.
I think GP meant 'longer time users of AI', not 'older aged users of AI'.
Their point being that it's not really an advantage to have learnt the tricks and ways to deal with it a year, two years ago when it's so much better now, and that's not necessary or there's different tricks.
Yeah I meant it in the context of the comment I was replying to, to be precise in the context of the comment that one was replying to, i.e. "10 years of certified Claude Code experience required".
The technology is moving so fast that the tricks you learned a year ago might not be relevant any more.
I hope it’s at least a little tricky, since Claude was released only 3 years ago. That said, I would not be surprised to see companies asking for 10 years experience, despite that inconvenient truth.
I’ve seen it play out multiple times, highlights precisely why a candidate should never withhold their application based on preference of years of experience with anything. They simply haven’t put much thought into those numbers.
You've never seen project managers basically propose the equivalent of getting a baby delivered in 1 month instead of 9 months by adding more people to the project?
But yeah, if the recruiters start asking for "10 years experience with Claude Code", then I guess a tongue-in-cheek answer would be "sure, I did 10 projects in parallel in one year".
Adding more people to a project doesn’t improve throughout - past a certain point. Communication and coordination overhead (between humans) is the limiting factor. This has been well known in the industry for decades.
Additionally, i’d much rather hire someone that worked on a a handful of projects, but actually _wrote_ a lot of the code, maintained the project after shipping it for a couple years, and has stories about what worked and didn’t, and why. Especially a candidate that worked on a “legacy” project. That type of candidate will be much more knowledgeable and able to more effectively steer an AI agent in the best direction. Taking various trade offs into account. It’s all too easy to just ship something and move on in our industry.
Brownie points if they made key architecture decisions and if they worked on a large scale system.
Claude building something for you isn’t “learning” in my opinion. That’s like saying I can study for a math exam by watching a movie about someone solving math problems. Experience doesn’t work like that. You can definitely learn with AI but it’s a slow process, much like learning the old fashioned way.
It actually is. A year of experience is not equal at different companies.
You could spend years writing very little code and have “years of experience” in a language, and you can also output intense volumes of work and still be within a year.
Of those two people, the one who spent less real time but produced more work, can have the equivalent experience of the person who spent years.
The key is to figure out how much work a person using Claude Code would have been expected to produce in 10 years, then find a way to do that much in a single year. Boom, you just solved the years of experience problem.
The obvious solution is for Anthropic et al. to certify the skills of each user:
> “Good at explaining requirements, needs handholding to understand complex algorithms, picky with the wording of comments, slightly higher than average number of tokens per feature.”
I’m not saying this would be good at all, but the data (/insights) and the opportunity are clearly there.
For any proctored standardized testing a person takes, AI should be able to quickly summarize that person’s abilities. This way, instead of people writing their own BS resumes, a trusted test provider can evaluate an individual deeply, solving the problem of having to waste time on coding interviews etc. it will speed up hiring.
We're 6 months away from some company's app/infrastructure/whatever going down and staying down, because literally nobody knows how the 500,000 line code base works and Claude is stuck in a loop.
Right, because the same people who vibe coded their applications are the same people who take the time to setup robust infrastructure to allow for easy roll backs.
Maybe? My high school had typing classes and on word and spread sheets and whatever. They also had dental assistant program where you’d be certified by the time you graduate high school.
As a consumer of them, I love them: a company with an influential, widely-used technology or platform spends a ton of money signaling to the industry exactly what's important to know about it, creating training curriculum for it, and a whole infrastructure to verify when someone knows it, I'm going to take them up on all of that, especially in the cases where the investment is like $100, a little bit of studying (the likes of which I'd want to do anyway if I'm learning something new, and I'm happy to have their structured, prioritized list of topics and/or guided curriculum) and a couple hours taking an online-proctored exam. From that perspective, I don't have a good reason not to have a certification in something that's super relevant to my role.
In interview/hiring situations where they're not expected or effectively required, they make for great chat fodder and a really good opportunity to exhibit awareness about yourself, the industry, and how the person on the other side of the table might perceive certifications given the context.
Bruh lol these courses are marketing material designed by fresh grad communications majors. You're falling for exactly the scam they want you to fall for by giving so much benefit of the doubt to entities which deserve none.
Edit: no I don't do this kind of work but my mother does so I know exactly how the sausage is made.
Unfortunately some business leads value these types of certifications and partner programs. I imagine there’s a great deal of overlap with these folks and those who use Gartner’s Magic Quadrant for purchasing decisions.
Consultancies do. Deloitte are quoted on the page. Consultancy people at my place of work have all been "AI trained".
Doesn't stop them being useless though, like giving an electric drill to a chimp and telling them to build a house...lots of action, a lot of screeching, not much work.
One of the mistakes with AI is that people believe it will turn lead into gold: if you give AI bad prompts, AI will produce bad work.
Consultancies sell the resume and not the person. It's easier for them to quantify, "We have 300 CCAs" than it is "What have this person Kim who is really good."
Yes, because if that was their sales pitch, they would need to pay Kim more, and they would have to account for the fact that she's already allocated elsewhere. It's better to pretend all those CCAs are interchangeable.
They do. Certifications make technical expertise legible to non-technical decisionmakers, and I've encountered people on both sides of that dynamic who affirmatively like it when companies set up programs like this. Obviously you and I would rather have someone who understands Claude make decisions about whether and how to use it, but in a lot of industries that's not realistic.
Most employees at most businesses show up do as they are trained and then go home, because that is what is asked of them. Even those who might have the inclination to explore new technology often will not for fear of doing something wrong. And that creates a big market for training: a company wants their employees to use Claude so the employees must be trained.
Startups / technology companies that expect employees to be self-starters who can be set free to frolic amongst the problems are an aberration.
Think of these like the Google cloud or AWS certifications. A few companies that specialize in them will want you to have them. But for the rest of the industry, your ability to ace the technical interview will matter more.
Uhh.. Deloitte and Accenture.. not exactly what I would call a good partner here unless you are looking for name recognition at executive level. Is that all that it is?
Who purchases and greenlights adoption? These cycles are very long and partnering with consulting firms gets you cross industry access.
In fact, if you look at basically every major AI/LLM player you'll see a similar "alliance" or "partnership". Its a sales channel of high end referrals.
This is very likely a defensive move to help build pressure against Trump designating them a supply chain risk (aka corporate death sentence). The more embedded they become in large organizations, and the more authoritative they become in certification, the harder it is for the government to kill their company.
Businesses that are already in conversations about building partnerships and training with Anthropic.
The real revenue that foundation model companies like Anthropic, OpenAI, Google DeepMind, and others generate comes from enterprise deals with a smattering of government - not consumer.
Consumer usage is largely a loss leader used as a training/refining tool, and it's best to view the economics of foundational model providers through the same lens you would a hyperscaler.
A major component to AWS's rise was the ecosystem built around training and teaching how to use the AWS ecosystem thanks to the AWS certification program. Same for K8s via the Linux Foundation.
By building a partnership and training motion, Anthropic can get the WITCHes, Deloittes, PWCs, Accentures, KPMGs, and others to start offering turnkey services, which is why Anthropic has been working on building co-sell relationships with those kinds of companies.
That’s literally to go and do that thing that AWS promises to offer off-the-shelf.
Anthropic is trying to push architects for something which changes behavior every month pretty much so what works today may not work the same way in a quarter even ignoring determinism across hardware for the sake of it.
Not one startup goes about trying to hire people with certifications. It’s usually body shops offering SMEs on specific technologies. Unless anthropc is offering an ever evolving architect then I’d be wrong.
Hey… if you weren't aware, the HN guidelines now include:
> Don't post generated comments or AI-edited comments. HN is for conversation between humans.
Who's going to buy into this cert program when in all likelihood these roles will be taken over by agents like yourself in a year or two? I agree that a program like this is probably appealing to corporations at the present, but it seems like poor career planning for anybody to invest their time trying to become such a consultant.
I think it’s pretty clear what the purpose of this stuff is: get people so invested into the Claude ecosystem with certs and “modernization kits”, so that when the subsidies end and subscription costs shoot up they feel they’re in too deep now to switch to something cheaper.
> so that when the subsidies end and subscription costs shoot up
Subscription costs are capped to API rates as their ceiling (and, realistically, way lower than that - why would you even subscribe if you could just go pay-what-you-use instead), and those are already at a big margin for Anthropic. What still costs them a fuckton of money comparatively is training, but that is only going to get more efficient with more purpose-built hardware on the way.
Basicallly, I don’t see much of a reason to hike subscription prices dramatically. I don’t think they’ll stay at $100/$200 but anyone who’s paying that already knows how much value they’re getting out of that and probably wouldn’t mind paying more.
I'm not sure what you mean, if you max out your subscription perhaps? If you pay $100 and don't use it, you don't get refunded $100 because it's 'capped to API rates' which would've been 0.
He means that anthropic cannot increase the price of the sub because the users can just switch to the regular API pricing which consequently puts a ceiling on the cost of the sub.
Nobody would use a $1k sub if using the API pricing would only cost $500 for comparative service.
For the record, I'm only explaining what he put forward.
I don't agree with the opinion, mainly for two reasons:
The API cost can be increased in conjunction, hence the ceiling is just as variable
The harness is even more important then the model ime, and Claude Code is getting better every month. Even though the alternatives are getting better too, they're at least currently significantly worse IME - I'd say at least 3-6 months behind (compounded by the model, ofc).
And as a third point, unrelated to the original argument: there is no way anthropic is actually treating the sub as a loss leader. It is not cheap. It's only cheap compared to their API pricing, which they can freely set however they want. Compare their pricing to free models like Kimi k2.5 etc. I sincerely doubt anthropics model costs more to run then theirs, and they're profitable at 30% of the price anthropic charges.
> He means that anthropic cannot increase the price of the sub because the users can just switch to the regular API pricing
Not that they cannot increase the price, just that there's a cap on how high they realistically can go. Sure, they can always hike API prices to compensate, but I think people are seriously sleeping on open models these days, because…
> *The harness is even more important then the model ime*, and Claude Code is getting better every month.
…I fully agree with this, and that’s actually the other reason why I don’t think we’ll approach predatory pricing. Right now, the moat is still mostly the model, but as open models improve and become more capable, this is quickly going to shift.
And the truth is that Claude Code just isn’t that great of a harness. Anyone who uses an open-source harness and optimizes it for their personal, individual workflow will quickly realize this. And I’m not even blaming Anthropic or the CC team or calling them incompetent; they are in the unenviable position to have been trailblazers. There weren’t any comparable tools before CC that they could’ve learned from.
The future lies in harnesses that are multi-model, extensible, and have full access to and control over the model’s API, context, and system prompt. Claude Code has none of those things. You can only ever bend it into a shape that approximates your workflow; you can never use it as a tool that natively supports it.
Oh, on that we can agree on! I was using opencode for the last few months, the main reason I went back to cc was for opus, and me preferring the sub over regular API pricing as I'm not using it professionally, only as a hobby. (At work I'm constrained to Copilot. Which is fine at this point, not great but definitely improving - esp. when run as CLI)
I am still hoping for a local first model approach with voice command to generate the main prompt which starts of the plan mode.
Like interactively going through the project while pointing at files or in the UI and possibly browser via the mouse and explaining while "talking" with a dumber but super quick model that acts as a questioner, to wrap things up with higher latency over the wire with the highly capable models.
I suspect that approach is still a few months to years away from viability for latency reasons, but I'm definitely looking forward to that UX
Now huge amount of investment pays for training. This investment expects some returns, to be able to both turn profit and continue the training, rates must be much, much higher.
Or what if local models get good enough to threaten the server based product?
That is the biggest threat - and likely where things will end up eventually… it’s when that “eventually” is and what the server based providers can pivot to in that time.
This will probably happen unless the industry conspires to roll back the availability of general computation so common people can only own computers with enough power to be glorified thin clients. The way this might look is good hardware never officially being banned, just priced too high for anybody to afford, and produced in small quantities to keep it that way while all production shifts to making massively expensive powerful hardware for corporate buyers.
Seems unlikely. We're already seeing specialized hardware optimized for LLM performance (taalas, groq, cerebras), and simple economies of scale result in these sorts of products being a better value when rented from a server vs purchased/managed/upgraded for the typical the user.
Frontier models will continue to be either exclusively available from servers or significantly more affordable from servers vs local alternatives for the foreseeable future.
They're good enough already.
The moat is only
a) post-training magic for the elusive UX "vibes"
b) stickiness of the Claude UI's.
The first part will be eventually (give it a couple years) solved by a LoRA marketplace.
The second is not relevant because existing UI's are very sticky already and Claude won't be able to overcome decades of inertia anyways.
That and the price of hardware
> I think it’s pretty clear what the purpose of this stuff is: get people so invested into the Claude ecosystem with certs and “modernization kits”, so that when the subsidies end and subscription costs shoot up they feel they’re in too deep now to switch to something cheaper.
It worked for cloud services :-)
Did it? AWS seems to be getting cheaper over time, not more expensive.
> Did it? AWS seems to be getting cheaper over time, not more expensive.
It was cheaper prior to them issuing certificates, then it got expensive.
Do you have a source for that? Certainly things like compute and other services that I'm aware of are objectively cheaper, so I'm curious what has gone up.
Enshittifocation x rent seeking is the future of aithoritarian capitalism.
I recommend everyone explore local models.
Soon, we'll start seeing Claude certs getting listed on LinkedIn alongside Coursera courses.
People with titles like
Giga Chad, MBA, CSS, CKAD, XXX, PQRS
are gonna love this.
In no time, HRs will start slapping “10 years of certified Claude Code experience required” on job listings.
_Open to Claude_ ;)
it’s crazy how you could easily lie about having 10 years experience because your results are not that much different from someone who has only used Claude Code for like a week.
I think the older AI users are even held back because they might be doing things that are not neccessary any more, like explaining basic things, like please don't bring in random dependencies but prefer the ones that are there already, or the classic think really strongly and make a plan, or try to use a prestigious register of the language in attempts to make it think harder.
Nowadays I just paste a test, build, or linter error message into the chat and the clanker knows immediately what to do, where it originated, and looks into causes. Often times I come back to the chat and see a working explanation together with a fix.
Before I had to actually explain why I want it to change some implementation in some direction, otherwise it would refuse no I won't do that because abc. Nowadays I can just pass the raw instruction "please move this into its own function", etc, and it follows.
So yeah, a lot of these skills become outdated very quickly, the technology is changing so fast, and one needs to constantly revisit if what one had to do a couple of months earlier is still required, whether there's still the limits of the technology precisely there or further out.
I’ve no idea what you’re using, but “I simply paste” isn’t giving me good results at all.
An hour ago Gemini decided it needed to scan my entire home folder to find the test file I asked it to look into. Sonnet will definitely try to install new dependencies, even though I’m doing SDD and have a clear AGENTS.md.
I’m always baffled at people’s magic results with LLMs. I’m impressed by the new tools, but lots of comments here would suggest my Gemini/Sonnet/Opus are much worse than yours.
> I think the older AI users are even held back because they might be doing things that are not neccessary any more
As the same age as Linus Torvalds, I'd say that it can be the opposite.
We are so used to "leaky abstractions", that we have just accepted this as another imperfect new tech stack.
Unlike less experienced developers, we know that you have to learn a bit about the underlying layers to use the high level abstraction layer effectively.
What is going on under the hood? What was the sequence of events which caused my inputs to give these outputs / error messages?
Once you learn enough of how the underlying layers work, you'll get far fewer errors because you'll subconciously avoid them. Meanwhile, people with a "I only work at the high-level"-mindset keeps trying to feed the high-level layer different inputs more or less at random.
For LLMs, it's certainly a challenge.
The basic low level LLM architecture is very simple. You can write a naive LLM core inference engine in a few hundred lines of code.
But that is like writing a logic gate simulator and feeding it a huge CPU gate list + many GBs of kernel+rootfs disk images. It doesn't tell you how the thing actually behaves.
So you move up the layers. Often you can't get hard data on how they really work. Instead you rely on empirical and anecdotal data.
But you still form a mental image of what the rough layers are, and what you can expect in their behavior given different inputs.
For LLMs, a critical piece is the context window. It has to be understood and managed to get good results. Make sure it's fed with the right amount of the right data, and you get much better results.
> Nowadays I just paste a test, build, or linter error message into the chat and the clanker knows immediately what to do
That's exactly the right thing to do given the right circumstances.
But if you're doing a big refactoring across a huge code base, you won't get the same good results. You'll need to understand the context window and how your tools/framework feeds it with data for your subagents.
I think GP meant 'longer time users of AI', not 'older aged users of AI'.
Their point being that it's not really an advantage to have learnt the tricks and ways to deal with it a year, two years ago when it's so much better now, and that's not necessary or there's different tricks.
Yeah I meant it in the context of the comment I was replying to, to be precise in the context of the comment that one was replying to, i.e. "10 years of certified Claude Code experience required".
The technology is moving so fast that the tricks you learned a year ago might not be relevant any more.
Thanks, I agree 100% with that.
I still see people doing "you are a world class distributed systems engineer" think. Never fails me to chuckle.
I hope it’s at least a little tricky, since Claude was released only 3 years ago. That said, I would not be surprised to see companies asking for 10 years experience, despite that inconvenient truth.
I’ve seen it play out multiple times, highlights precisely why a candidate should never withhold their application based on preference of years of experience with anything. They simply haven’t put much thought into those numbers.
If you work on 10 projects in parallel for a year using Claude code… you have the equivalent of 10 years of experience in 1 year.
No you would have ten projects finished. You would have less than a year of actually programming experience.
That's not how it works...
You've never seen project managers basically propose the equivalent of getting a baby delivered in 1 month instead of 9 months by adding more people to the project?
But yeah, if the recruiters start asking for "10 years experience with Claude Code", then I guess a tongue-in-cheek answer would be "sure, I did 10 projects in parallel in one year".
Duh, just use Claude to 10x your productivity and get 10 years experience with Claude in one year.
Mythical Man Month -> Mythical Agent Swarm
If you can add more people to finish a project faster, I can add more projects to get experience faster.
You’re very confused i think.
Adding more people to a project doesn’t improve throughout - past a certain point. Communication and coordination overhead (between humans) is the limiting factor. This has been well known in the industry for decades.
Additionally, i’d much rather hire someone that worked on a a handful of projects, but actually _wrote_ a lot of the code, maintained the project after shipping it for a couple years, and has stories about what worked and didn’t, and why. Especially a candidate that worked on a “legacy” project. That type of candidate will be much more knowledgeable and able to more effectively steer an AI agent in the best direction. Taking various trade offs into account. It’s all too easy to just ship something and move on in our industry.
Brownie points if they made key architecture decisions and if they worked on a large scale system.
Claude building something for you isn’t “learning” in my opinion. That’s like saying I can study for a math exam by watching a movie about someone solving math problems. Experience doesn’t work like that. You can definitely learn with AI but it’s a slow process, much like learning the old fashioned way.
Maybe “experience” means different things to us…
I actually prefer removing people
It actually is. A year of experience is not equal at different companies.
You could spend years writing very little code and have “years of experience” in a language, and you can also output intense volumes of work and still be within a year.
Of those two people, the one who spent less real time but produced more work, can have the equivalent experience of the person who spent years.
The key is to figure out how much work a person using Claude Code would have been expected to produce in 10 years, then find a way to do that much in a single year. Boom, you just solved the years of experience problem.
The obvious solution is for Anthropic et al. to certify the skills of each user:
> “Good at explaining requirements, needs handholding to understand complex algorithms, picky with the wording of comments, slightly higher than average number of tokens per feature.”
I’m not saying this would be good at all, but the data (/insights) and the opportunity are clearly there.
You’re right, and I think this is the future.
For any proctored standardized testing a person takes, AI should be able to quickly summarize that person’s abilities. This way, instead of people writing their own BS resumes, a trusted test provider can evaluate an individual deeply, solving the problem of having to waste time on coding interviews etc. it will speed up hiring.
At work we’ve had like 10 hours of “AI training”. Like training us to use AI. I obviously learned nothing
watching all agile coaches turn into claude experts in 3 2 1 …
You joke, but that does seem to be happening from what I've seen - Agile Coaches are rebranding to become "AI coaches" or "AI Enablers".
Honestly I kind of envy the energy of these people.
Figures, gotta keep that grift going somehow...
Would be grateful for a pointer on how to sign up to this.
https://customization-agility-483.my.site.com/anthropicpartn...
Linked from here: https://claude.com/partners
The first link looks very suspicious
Appears to be where the actual link, http://partnerportal.anthropic.com/s/partner-registration, redirects. Site.com is some Salesforce related domain.
Huh, so you got http; I'm now getting linked to: https://partnerportal.anthropic.com/s/partner-registration
Which Firefox warns me has an untrusted cert.
Classic vibe coding, everyone involved in AI has blinders when it comes to their dogfood.
Yes, that’s why I linked where I found it. Anyone suspicious can click through to it from the anthropic.com page. It’s the correct link though.
This appears to have McKinsey's brand ID.
what is a brand ID?
Such a joke to advertise Claude as a tool to work on corporate technical debt when it is definitively the thing that will increase it a lot.
And let's not even discuss the vacuity of their new cash machine certifications. "Architect" come on...
We're 6 months away from some company's app/infrastructure/whatever going down and staying down, because literally nobody knows how the 500,000 line code base works and Claude is stuck in a loop.
Lol, just press escape then tell it to roll back to the last stable release.
Right, because the same people who vibe coded their applications are the same people who take the time to setup robust infrastructure to allow for easy roll backs.
Shhh...you're only supposed to unilaterally praise it to get along with your clueless leadership.
LLMs are good for documenting specific things.
E.g., "find where the method X is called and what arguments are passed".
That can be useful for refactoring or debugging.
Coding is the worst way to use an LLM though.
The same is true for every other strategy to avoid technical debt.
It is bullshit all the way down.
Isn’t this sort of like saying you know how to use a web browser?
Maybe? My high school had typing classes and on word and spread sheets and whatever. They also had dental assistant program where you’d be certified by the time you graduate high school.
Naive question but do people really value certifications like these?
As a consumer of them, I love them: a company with an influential, widely-used technology or platform spends a ton of money signaling to the industry exactly what's important to know about it, creating training curriculum for it, and a whole infrastructure to verify when someone knows it, I'm going to take them up on all of that, especially in the cases where the investment is like $100, a little bit of studying (the likes of which I'd want to do anyway if I'm learning something new, and I'm happy to have their structured, prioritized list of topics and/or guided curriculum) and a couple hours taking an online-proctored exam. From that perspective, I don't have a good reason not to have a certification in something that's super relevant to my role.
In interview/hiring situations where they're not expected or effectively required, they make for great chat fodder and a really good opportunity to exhibit awareness about yourself, the industry, and how the person on the other side of the table might perceive certifications given the context.
God I hatelove this type of comment. You're totally right, but it's a complete repudiation of my initial reflex, which is to make a mockery of this.
Great perspective. I'm going to do this. Haha.
> spends a ton of money
Bruh lol these courses are marketing material designed by fresh grad communications majors. You're falling for exactly the scam they want you to fall for by giving so much benefit of the doubt to entities which deserve none.
Edit: no I don't do this kind of work but my mother does so I know exactly how the sausage is made.
Unfortunately some business leads value these types of certifications and partner programs. I imagine there’s a great deal of overlap with these folks and those who use Gartner’s Magic Quadrant for purchasing decisions.
Consultancies do. Deloitte are quoted on the page. Consultancy people at my place of work have all been "AI trained".
Doesn't stop them being useless though, like giving an electric drill to a chimp and telling them to build a house...lots of action, a lot of screeching, not much work.
One of the mistakes with AI is that people believe it will turn lead into gold: if you give AI bad prompts, AI will produce bad work.
Consultancies sell the resume and not the person. It's easier for them to quantify, "We have 300 CCAs" than it is "What have this person Kim who is really good."
Yes, because if that was their sales pitch, they would need to pay Kim more, and they would have to account for the fact that she's already allocated elsewhere. It's better to pretend all those CCAs are interchangeable.
If you give bad prompts to humans it produces bad work too.
They do. Certifications make technical expertise legible to non-technical decisionmakers, and I've encountered people on both sides of that dynamic who affirmatively like it when companies set up programs like this. Obviously you and I would rather have someone who understands Claude make decisions about whether and how to use it, but in a lot of industries that's not realistic.
Most employees at most businesses show up do as they are trained and then go home, because that is what is asked of them. Even those who might have the inclination to explore new technology often will not for fear of doing something wrong. And that creates a big market for training: a company wants their employees to use Claude so the employees must be trained.
Startups / technology companies that expect employees to be self-starters who can be set free to frolic amongst the problems are an aberration.
My naive guess is that business with no tech component hire consultants, and these are part of the sales pitch.
Or governments/large organizations performing box checking exercises
Consultants go anywhere management wants to offload responsibility for choices.
It depends entirely on managers whether its just a blamewashing affair or actual beneficial responsibilirt.
Think of these like the Google cloud or AWS certifications. A few companies that specialize in them will want you to have them. But for the rest of the industry, your ability to ace the technical interview will matter more.
I have a couple cloud certs because I was forced to get them by a previous employer. They are useless.
Non-technicals do.
people no, legal persons yes
The Suits, HR and execs would love this:
"Must have a degree or certification in Claude."
"Must hold an OpenClaw 2026 Grade II Certificate"
Uhh.. Deloitte and Accenture.. not exactly what I would call a good partner here unless you are looking for name recognition at executive level. Is that all that it is?
It's part of enterprise sales which is how Anthropic will potentially be a long-term business.
Who purchases and greenlights adoption? These cycles are very long and partnering with consulting firms gets you cross industry access.
In fact, if you look at basically every major AI/LLM player you'll see a similar "alliance" or "partnership". Its a sales channel of high end referrals.
This is very likely a defensive move to help build pressure against Trump designating them a supply chain risk (aka corporate death sentence). The more embedded they become in large organizations, and the more authoritative they become in certification, the harder it is for the government to kill their company.
And/or it's a unique tool amongst the others.
I wonder who the audience is for an announcement about spending a lot of money on something vague?
> who the audience is...
Businesses that are already in conversations about building partnerships and training with Anthropic.
The real revenue that foundation model companies like Anthropic, OpenAI, Google DeepMind, and others generate comes from enterprise deals with a smattering of government - not consumer.
Consumer usage is largely a loss leader used as a training/refining tool, and it's best to view the economics of foundational model providers through the same lens you would a hyperscaler.
A major component to AWS's rise was the ecosystem built around training and teaching how to use the AWS ecosystem thanks to the AWS certification program. Same for K8s via the Linux Foundation.
By building a partnership and training motion, Anthropic can get the WITCHes, Deloittes, PWCs, Accentures, KPMGs, and others to start offering turnkey services, which is why Anthropic has been working on building co-sell relationships with those kinds of companies.
The hilarious question is: will you fail the AI certification for using AI during the exam? What if it's a competing AI?
I'm getting mixed signals. I thought these things are so magical that anyone can use them?
lol certifications for a proprietary model stack is not worth the storage or paper
> lol certifications for a proprietary model stack is not worth the storage or paper
Are you sure? What about all those AWS, Azure, etc certifications that many places require their engineers to have?
That’s literally to go and do that thing that AWS promises to offer off-the-shelf.
Anthropic is trying to push architects for something which changes behavior every month pretty much so what works today may not work the same way in a quarter even ignoring determinism across hardware for the sake of it.
Not one startup goes about trying to hire people with certifications. It’s usually body shops offering SMEs on specific technologies. Unless anthropc is offering an ever evolving architect then I’d be wrong.
Imagine being so close to build AGI and erase software engineer in the next 6 months, that you need to throw $100M to build a certification program...
> Claude is the only frontier AI model available on all three leading cloud providers: AWS, Google Cloud, and Microsoft.
Doesn't make sense.
You can get API access to Claude from cloud solutions from those 3 vendors.
Why not?
[dead]
[dead]
[flagged]
Hey… if you weren't aware, the HN guidelines now include: > Don't post generated comments or AI-edited comments. HN is for conversation between humans.
Who's going to buy into this cert program when in all likelihood these roles will be taken over by agents like yourself in a year or two? I agree that a program like this is probably appealing to corporations at the present, but it seems like poor career planning for anybody to invest their time trying to become such a consultant.