Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Stop selling “unlimited”, when you mean “until we change our minds”

blog.kilocode.ai

Post reply on HN
Print view View original

311–320 of 363 posts

Re: Stop selling “unlimited”, when you mean “until we change our minds”

#311

Post by thimabi »

vertoc wrote: post #117
thimabi wrote: post #63

Earlier quoted context omitted.

Things would be much easier if, instead of “unlimited”, companies simply offered great usage limits. Yes, hoarders would still exist, but their impact would be negligible. Even better: provide a counter displaying both remaining usage available and the quota reset time. But companies probably earn so much money from the vast majority of users that having good and clear limits would only empower them to actually benef…

I think the hard part with these ai models is it’s kind of hard to figure out how many tokens you’re going to use, especially as a new user. 4,000 tokens sounds like a lot for instance, but is tiny

If companies offered any sort of counter, users would quickly understand their usage patterns and adapt accordingly. But users are kept in the dark on purpose, often not knowing how many tokens they use, and sometimes not even knowing what are the token limits.

Re: Stop selling “unlimited”, when you mean “until we change our minds”

#312

Post by nottorp »

philipp-gayret wrote: post #269
nottorp wrote: post #252

Earlier quoted context omitted.

Think larger codebases that do not involve node/npm...

The reason I ask is I work on developer experience, and I see feedback often online by developers that AI simply doesn't work for their flow. So I'm looking for real-world concrete examples on what AI development tools are stuggling with. Or, maybe it is how you use the tool? Personally I haven't run into limitations, so I'm really looking for hard counter-examples. The Godot example by the other poster was great, ma…

Can't help because I haven't tried those bots that edit the code for you.

I just use "AI" instead of Google/SO when I need to find something out.

So far it mostly answers correctly, until the truthful answer comes close to "you can't do that". Then it goes off the rails and makes up shit. As a bonus, it seems to confuse related but less popular topics and mixes them up. Specific example, it mixes couchdb and couchbase when I ask about features.

The worst part is 'correctly' means 'it will work but it will be tutorial level crap'. Sometimes that's okay, sometimes it isn't.

So it's not that it doesn't work for my flow, it's that I can't trust it without verifying everything so what flow?

Edit: there's a codebase that i would love to try an "AI" on... if i wouldn't have to send my customer's code to $random_server with $random_security with $untrustable_promises_of_privacy. Considering how these "AI"s have been trained, I'm sure any promise that my code stays private is worth less than used toilet paper.

Gut feeling is the "AI" would be useless because it's a not-invented-here codebase with no discussion on StackOverflow.

Re: Stop selling “unlimited”, when you mean “until we change our minds”

#313

Post by raverbashing »

wat10000 wrote: post #159
raverbashing wrote: post #62

Earlier quoted context omitted.

Still limited. But "unlimited" to most people In the same way your next-door supermarket has effectively "infinite soup cans" for the needs of most people.

Unlimited means for a flat rate. Pay per use/item, like at the supermarket, isn’t it.

Still, you can't get "unlimited" soup cans from your next door supermarket even paying

Re: Stop selling “unlimited”, when you mean “until we change our minds”

#314

Post by jwatte »

Bluestein wrote: post #128
epistasis wrote: post #54

Earlier quoted context omitted.

I had been loyal, but am not any longer, so the ploy definitely did not work with me. I guess I'll move on to Gemini now, until I get sick of it. To the degree there is a moat, I do not think it will be effective at keeping people in. I had already been somewhat disillusioned with the AI hype, but now I am also disillusioned with the company who I thought was the best actor in the space. I am happy that there is unli…

> I had already been somewhat disillusioned with the AI hype, but now I am also disillusioned with the company Likewise: a faulty, unproven, hallucinating, error-prone service, however good, was a good value at approx 25 USD/month in an "absolutely all you can eat", wholesale regime ... ... now? Reputational risk aside, they force their users to appraise their offering in terms of actual value offered, in the market.…

> they force their users to appraise their offering in terms of actual value offered

That's a good thing, right?

Re: Stop selling “unlimited”, when you mean “until we change our minds”

#315

Post by Bluestein »

jwatte wrote: post #314
Bluestein wrote: post #128

Earlier quoted context omitted.

> I had already been somewhat disillusioned with the AI hype, but now I am also disillusioned with the company Likewise: a faulty, unproven, hallucinating, error-prone service, however good, was a good value at approx 25 USD/month in an "absolutely all you can eat", wholesale regime ... ... now? Reputational risk aside, they force their users to appraise their offering in terms of actual value offered, in the market.…

> they force their users to appraise their offering in terms of actual value offered That's a good thing, right?

Very much, absolutely, most resolutely and definitely, yes.-

Re: Stop selling “unlimited”, when you mean “until we change our minds”

#317

Post by Aeolun »

Aurornis wrote: post #249
Aeolun wrote: post #121

Earlier quoted context omitted.

> LLM subscriptions need to go away, why can’t we just pay as we go? It’s the fairest way for everyone. Because Claude Code is absolutely impossible to use without a subscription? I’m fine with being limited, but I’m not with having to pay more than $200/month Anybody that feels they’re not getting enough out of their subscription is welcome to use API instead.

> Because Claude Code is absolutely impossible to use without a subscription? Claude Code accepts an API key. You do not need a subscription https://docs.anthropic.com/en/docs/claude-code/settings#envi...

I mean the way it works quickly leads to exorbitant costs. Subscription based pricing with limits keeps everyone honest.

Re: Stop selling “unlimited”, when you mean “until we change our minds”

#318

Post by crashprone »

shlant wrote: post #227
apwell23 wrote: post #169

Earlier quoted context omitted.

so where did 'bad users' come from if users were simply doing what they were allowed to? why is anthropic tweeting about 'naughty users that ruined it for everyone' ?

are you going to answer what the bait and switch was?

Bait : "For 200$ a month you get to use Claude 20x more than what the Pro users are entitled to. You don't know how much exactly though, but neither do we. We may limit your usage with weekly and monthly limits. Sounds good?"

Switch: "We limited your usage weekly and monthly. You don't know how those limits were set, we do but that's not information you need to know. However instead of choosing to hoard your usage out of fear of hitting the dreaded limit again, you've kept it again and again, using the product exactly the way it was intended to and now look what you've done."

Re: Stop selling “unlimited”, when you mean “until we change our minds”

#319

Post by const_cast »

jollyllama wrote: post #174
thimabi wrote: post #63

Earlier quoted context omitted.

Things would be much easier if, instead of “unlimited”, companies simply offered great usage limits. Yes, hoarders would still exist, but their impact would be negligible. Even better: provide a counter displaying both remaining usage available and the quota reset time. But companies probably earn so much money from the vast majority of users that having good and clear limits would only empower them to actually benef…

It simply shouldn't be called unlimited, then.

This is the crux of the issue.

Every company wants the marketing of unlimited, but none of them want the accountability.

Re: Stop selling “unlimited”, when you mean “until we change our minds”

#320

Post by bubblyworld »

NoGravitas wrote: post #307
theshrike79 wrote: post #153

Earlier quoted context omitted.

Just completely ignoring LLMs is like not wanting to use autocomplete or language servers, preferring to hand-craft everything. But on the other hand there is the other end who think AGI coming in a few months and LLMs are omniscient knowledge machines. There is a sweet spot in the middle.

For me, the sweet spot seems to be "keep an eye on LLM developments, but don't waste work time trying to use them yet".

Yeah, that makes sense to me for a lot of reasons. Recently decide to take the plunge and start introducing AI tools in my startup. Fairly cheap experiment, all things considered - three months of giving it a try as a team and then we'll make a call. Time will tell!
Post reply on HN