gavanwilhite

This is thoughtful and should indeed go beyond AI risk (to bio, etc).

Employees should have protected ways of disclosing unsafe practices, particularly when the potential impacts are so large.

Fantastic context. I love hourglass shaped brainstorming, and the “questions” prompt is a solid one!

Transparency mode alone is a technological marvel.

It is extraordinarily difficult to do correctly, and once your brain has adapted to the Apple acoustic transformations, they’ve achieved near-indistinguishability.

The US is pretty keen to enforce its tax law internationally. If we did a wealth tax, there would be a number of options to limit using movement or citizenship renouncement as an easy out.

Something that would be extra awesome would be light markdown support inside fields. This would make it useful for reviewing chatbot conversations.

Splash screen removal, access to support, etc. Generally all the benefits that come with paying for a thing, rather than being on a tier that the company is trying to get you off of.

Killing the Plus plan is the worst part about this. The cheapest paid tier is now $2000 / year, putting it out of range for side projects.

It was Unity until today... They just canceled the Plus tier, making the cheapest paid plan $2000 / year.

While this sounds wise at first glance, this is just clearly not true to anyone who's used ChatGPT.

The problem is usually the opposite (sycophancy)

Does this have any ability to cache GraphQL requests, or is there a recommended strategy for doing so?

Seriously.. I just posted a rather nuanced question (can a CSS gradient extend into over-scroll like color can), and it just got instantly closed by someone who didn’t understand the question at all. Apparently I need “cred” to even request review on this using their meta channels. Really bad FTUE.

https://stackoverflow.com/questions/75240915/is-it-possible-...

I think this was done by the author to create engagement with their blog.

I'm not sure if there is a against this, but it feels kinda against the spirit of posting a title like this when it seems like the goal of the linked page is 20% to communicate the topic of the headline and 80% to drive you to other topics they have been posting about.

Looks promising. I tried signing up for your newsletter, but I don’t see a submit button and enter doesn’t work

I think that this has a way higher chance of being true than people have been assuming.

The slack convo leak would have been the same either way.

It really has a strong smell of protect-my-ego.

We’ve been using it to make quests :-)

https://questgiver.org/t/05c0a7d7-5f22-41d2-90aa-f02082fbb8d...

These were entirely generated based on an intro prompt + “Make three quests people could do to help ensure ethical AI development”. It generated a json with objective types, emojis, everything.

We’re planning to use it to get past the “blank page” problem when onboarding organizations.

Thank you for this comment!

I haven't spent much time learning about how ChatGPT works, but:

'ChatGPT does not simply output the “most semantically popular result”. That description applies only to the base model, before instruction tuning and RLHF.'

was exactly the explanation I needed!

Will keep in mind going forward. Thanks for the kind way you point out rules!

Some great news comes out of the WotC OGL controversy, with Paizo announcing a new open license and taking a strong legal stand against WotC.

They will be looking towards groups like the Linux Foundation to own the license once developed.

(Site has been overwhelmed with traffic, thus archive.org)

“Greedflation”, that’s a great term.

Too many folks using the media hubbub on inflation, economic ‘predictions’, etc as memetic cover for greed-based maneuvering.

Great framing! Easy visual to remember :-)

Yes, they do this! I called support and they agreed that it was in opposition to their website description of Apple Arcade, but that there was nothing they could do.

It pops up on the top of the screen every time you open the game.

They were very cagy about it, but their excuse seemed to be that it was up to the game publisher whether to include it, and that I should complain to them (even though it was sent as an Apple Arcade notification).

https://imgur.com/a/KOYSyTm

Woah, this female voice is clearly sampled from one of the Audm narrators.

Read to the end!

I’d like to politely point out that you are speaking to the geopolitical coordination challenges related to deployment, not the technology itself.

Those are real, important issues to be worked through, and to your point, and they are better served by a multi-stakeholder approach that values climate justice.

There is research and work needed both on policy and the technology aspects of this.

The technology is good here. The “move fast and break things” approach is not.

I know your intentions here are good.

However, you aren’t being “edgy” or “bold”, you are being unwise.

This is not a place to shoot from the hip.

Rationalizing from science fiction, has serious failure-modes, and your approach risks poisoning public support for an important climate intervention.

If you want to do right by this, you need to steelman the scientists critical of your work. Reflect back to them the best version of their argument until they agree that you understand it.

Also!

Many folks who smoke know the dangers, but don’t know how to quit, or fear being miserable.

I’d highly recommend offering a book from https://www.allencarr.com/

They might not read it for a few years, but when they do, it’s the best way to quit.