At the very least, Ilya Sutskever genuinely believed it, even when they were just making a DOTA bot, and not for hype purposes.

I know he's been out of OpenAI for a while, but if his thinking trickled down into the company's culture, which given his role and how long he was there I would say seems likely, I don't think it's all hype.

Grand delusion, perhaps.

I suppose this Sealed Sender issue is problematic for some people, but it's not enough for me to seriously consider jumping ship.

I couldn't understand how a 2022 device would run so much worse than 2017 device and assumed it was faulty. Returned, given a replacement, same issue. It is quite literally not built to hand the heat from the Intel chip doing very minimal stuff. I refuse to use a laptop that sounds like a jet engine when Microsoft is doing basic background stuff.

Returned and ended up buying a used 15 inch T-type Thinkpad with an AMD chip recommended by Reddit for $500 NZD. Runs great, cool, and quiet. It's much bigger and bulkier that the Dell but I don't mind.

Note: Not a Thinkpad fanboy, work has given me an X1 Carbon that I dislike for the same reasons I didn't like the new XPS 13 - it's useable, but it's still much hotter and louder than I would like.

Always sad to see projects like this go. This was probably the best alternative to GrapheneOS for non-Pixel devices.

It's an Obsidian knock-off. It's pretty janky and the documentation is lacking. It's open-source which is nice... But the company behind it is ??? I don't know. They are Chinese but I couldn't find much about them.

I use it because I can self-host it, it has most of the features Obsidian has, and I can use it in a web-browser from anywhere - which is the biggest feature for me that Obsidian lacks.

It really is, and I see more and more of it in Reddit comments, and even at work.

I had some obvious AI writing sent to me by a lawyer on the other side of a dispute recently and I was pissed - I don't mind if you want to use it to help you (I do myself), but at least have the decency to edit so it doesn't read like ChatGPT trash.

You take the video transcript, ask ChatGPT to write a short blogpost about it, and this is what you get.

Example: Previously, Pixel 6 was to receive 3 years of OS updates and 5 years of security updates. Now it will receive 5 years of both.

Always important to bear in mind that the examples they show are likely the best examples they were able to produce.

Many times over the past few years a new AI release has "wowed" me, but none of them resulted in any sudden overnight changes to the world as we know it.

VFX artists: You can sleep well tonight, just keep an eye on things!

Comments URL: https://news.ycombinator.com/item?id=41740951

Points: 4

# Comments: 0

She said no.

So they found a soundalike and Tweeted out references to the movie Her (starring Scarlett Johanssen as an AI chat bot) in the days leading up to launch.

Scummy as fuck from OpenAI regardless of the technical legal rights and issues involved.

The lack of E2EE is actually what has got Telegram in trouble.

The same stuff that happens on Telegram also happens on WhatsApp and Signal - but because it's E2EE, it's harder for the police to find it, and when they do find it, WhatsApp and Signal have no way to comply - they can't see the chats! That said, they will provide details to LE that they do have access to (e.g. metadata) - of which WhatsApp holds a lot more.

In Telegrams case with public groups sharing illegal content out in the open - this is viewable to LE and they request logs from Telegram for their investigations. This stuff is publicly viewable on and to Telegram. Telegram can remove it. Telegram can provide chat logs to LE - but they refuse to. This is why they're in trouble.

Moral of the story - if you want to start an internet service - you either go E2EE no knowledge, or you moderate your shit and cooperate with LE, you can't avoid doing the latter unless you're doing the former.