timcobb

Checks out... the Internet is the thing that more humans are working everyday to change than anything else on the planet.

I wonder what people think about addressing this with process. For example, on GH:

- Disable public pull requests.

- Require people to open an issue or discussion. Issues and discussions should have stated length/quality parameters. If an issue is a wall of Claude-text, users should be prepared for this text to be automatically summarized into plain language. If you don't want your Claude-text to be machine-turned into something human-consumable, the onus is on you to post human text up front.

- PRs are only drafted once approved in issue or discussion

I'm wondering if this is a tractable approach that yields results. I've seen references to a few projects trying something like this. Would be nice to hear folks experience.

> This is why there was a "Sonnet only" usage bar for Max tier for the longest time.

it's still there. I still don't totally grok why I can't use all my tokens on Sonnet if I want to... maybe that signals something?

The question is whether the m-dashes are surrounded by spaces or not. The spaces are utterly maddening. But yeah, RIP the mdash, who would have thought.

Do you write /maintain evals? This is something I want to get into more. Otherwise I feel really blind and feel compelled to just drop money on frontier.

What provider(s) do you use?

Thanks. I was looking at open code go yesterday and I couldn't figure out if the base pricing is including usage or if that's just base pricing and then you have to pay for usage too. How does it work? It is very cheap.

Can people share their GLM and open model setups in general please? What provider do you use. Why do you trust it with serving full quality? What harness do you use? Why do you trust it not to have malware (most harnessed are TS apps). I am just trying GLM 5.1 from Nvidia build in open code would love to hear how you all do it, thanks.

/rc has improved my health

You can vote against this trash and still not consume 99% of the "news" content people pretend is information. You don't need to know "the latest" about this second "administration". Consuming "news" is not a civic duty, it just is bad for you and a waste of your time. Don't be fooled.

I live in Baltimore county and most of our local news is homicides and stuff like that.

I know some people who moved to Delaware beaches and they told me that there the local news is just folks talking about government administration stuff, which was pretty mind-blowing to me. I would like to check this out.

Deaf people have a lifetime of training (hopefully) and experience dealing with things like this in ways the are appropriate for their situation

Big talk from US consumers. The reality is we'll consume those ads and we'll love it. Sir, yes sir!

Not in my subjective experience sadly

this sounds a bit different than a university library situation

I think pooling/donating tokens will be a thing. Not sure if like this, but in some format. The Django project, for example, came out and said they don't want your tokens, but I think a lot of people/projects will (do?) want your tokens.

This is fitting/perfect. Pokemon go is THE archetypical surveillance capitalism app. Be a drone in surveillance capitalism, know that your behavior will be used like this. Drones generally don't know or care though. Drones just have fun with tech yay fun awesome. Pokemon, gotta catch them all!!

Damn

But I don't want to be the developer who goes and says we must pay all this money for these tokens. I don't know who wants to be that developer.

That make sense, but what about the specific bifurcation we're seeing here of super primo models versus still good models being available to subscriptions?

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??

I did, I'm not seeing anything about the future of subscriptions at Athropic.

What's their word? Have they commented?

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?

Ooof so are we thinking that in the next 6-12 months subscriptions will be replaced with paying retail like enterprise currently?

It's interesting living through all this civilizational flailing. Remarkable. I read that when people invented writing it was very controversial because the facility of memory was the prized intellectual capability of the time. I never thought I'd see this kind of thing in person

Yeah this is incredibly high quality (at least on the surface, who knows what's underneath). Great job authors.

This is going to be very different though. These companies will just start investing in in house cheaper inference that's been on the front page here the last few weeks.

You can also ask Claude and get an immediate answer, the power is yours

Why give anyone benefit of doubt when you're just trying to extract true information. If a person BSs you, you don't believe them in the future either. Giving BS the benefit of doubt isn't serving anyone. Giving intent benefit of doubt makes sense, but intent isn't important when it comes to technical information and lying about it

Why does this make it better for you the consumer? You still need to detect the lie and respond to it appropriately. Machines lie more elegantly, I think is the scary part

How different is this from humans? In my experience people fill in the blanks quite the same, they just do it less convincingly and sometimes maliciously. Having the sort of prejudice you describe against AI content doesn't make sense if you consider humans make mistakes and lie all the time, coming from a position of less knowledge than LLMs. You need to approach both with similar caution.