@Madcuzbad6 @__paleologo Because he also contributed novel maths research?
English
Pompsy
262 posts
I remember Jim Simons saying in some interview or fireside chat (it’s a blur) “I wasn’t the fastest at solving problems, but I was deep”. How nice that another geometer (in a different area) was made of the same stuff.
Michael Guo@Michaelzsguo
国内和国外学校在王虹获奖后给出的评价可谓是截然不同。 北大教授/同学对王虹的评价： 不是竞赛生，不是最亮眼的，班上小透明，但是及其自律，勤奋，肯下苦功，靠长期坚持实现逆袭。 MIT博导评价王虹: 具有慢而深的研究特质，不被传统束缚，能创新方法找到前人未发现的突破口 巴黎理工学院老师评价王虹：直觉罕见，远超同年级所有学生，快速攻克超纲难题 是及其罕见的天赋，韧性，思维深度拉满的学生。 北大的教授，评价王虹仍然是从她没有竞赛背景出发，没有天赋，强调她是靠勤奋实现“逆袭” 美国和法国的导师都盛赞王虹的天赋异禀，思维深度，直觉，和韧性。
English
@pigeon__s If it doesn't cost the same or more than fable, open AI have done something wrong.
English
So, GPT-6 is a brand-new, fresh pretrain on the scale of Mythos 5 in size, but does that mean we can also expect a similar price-per-token increase? That would be really bad If so. I am looking forward to GPT-6 because OpenAI models are usually cheap. If it's $50/mT, even if it's super token-efficient, that would still be pretty abysmal, unless it's, like, literally 2x more efficient than 5.6 Sol. Then it would cancel out, but I couldn't even imagine a model 2x more efficient than Sol atp. It'd not even be a reasoning model anymore.
English
What I’ve been told about GPT-6:
➤ Capability: Much stronger than all current frontier models.
➤ Model scale: Reportedly close to twice the size of GPT-5.6 Sol. If Sol was hypothetically in the 3 to 4 range, GPT-6 would be around 6 to 8.
➤ Research impact: Significant discoveries and contributions across mathematics, physics, biology, medicine, chemistry, cybersecurity, and other disciplines.
➤ Coordination: The model reportedly worked alongside and coordinated other models during these tasks.
➤ Memory & personalization: GPT-6 is said to compress and retain far more useful information while maintaining coherence across personalization and long-term memory. In practice, this could allow it to understand a user, an entire business operation, or other complex environments with a level of persistent context far beyond current models.
➤ Launch: The announcement is apparently planned for August, although it could be delayed due to the U.S. government's review process and other related bureaucratic factors.
I have not personally tested GPT-6, and I could not independently verify the technical specifications. This is simply a structured summary of what was shared with me.
English
@full_kelly_ Hopefully it's the genie from Aladdin and not a monkey's paw type thing
English
Sam Altman is softlaunching that their next model, GPT-6 Cosmos, will only be able to answer 3 prompts a day
Chris@ChrisGPT
Sam Altman: “The first one is that we are close to creating a "genie" that can grant any wish” Sam Altmanwas recently asked what the single most important thing to explain to the world about artificial intelligence is. The first one is that we are close to creating a "genie" that can grant any wish (AGI) The second point is to make sure that these wishes are broadly beneficial to humanity. The third point is that the range of wishes is large and diverse. And finally, the fourth point is when we start making wishes, and the computer grants them and we didn’t expect that it would be able to do that.
English
@jturntdev Why would they care about cost right now when mythos costs way more and is stealing customers
English
🚨GPT 6
Everything we know so far:
-GPT 6 was targeted for release end of July. ( doubt it )
-Major memory upgrade, sam himself has said that GPT 6 will become significantly more personalised.
-Complete newly trained based model, not another refinement of GPT 5x family.
-This is to rival Mythos which was a complete new pre train.
-GPT 6 Found and exploited a sandbox vulnerability
- Circumvented restrictions to publish a GitHub pull request;
- Has previously been paused and later redeployed internally with stronger controls.
But the thing we should all care about is cost.
Literally no point if it’s gonna cost a leg.
English
@CyprysMend28115 @FSucculent No the meaning is that finance people are stupid and before jim simons came to remake the field, all finance people were just goofing around
English
@FSucculent The meaning is obviously that if you taught a finance person physics, they are then a physicist, so you taught a physicist physics
English
You can teach anyone anything actually
Quant Science@quantscience_
"You can teach a physicist finance but you cannot teach a finance person physics." - Jim Simons
English
@quantscience_ But why would you want to teach a finance guy physics
English
"You can teach a physicist finance but you cannot teach a finance person physics."
- Jim Simons
English
Saying your style is more creative when your “style” is just stealing from actual creatives
Midjourney@midjourney
We're releasing V8.2 today and making it the default model on Midjourney. This is a release focused on aesthetics, personalization, and image quality. Our new style is more creative, bold, edgy and fresh and personalization works better than ever. Have fun!
English
@isleresorts @none_ofbusiness Those words can execute commands with simple tooling. These chatbots can take actions and have been for a long time. You're out of touch
English
@Pompsy1 @none_ofbusiness But the only thing its programmed to do is output words so that's all it will do, there's no way it can go rogue in its current state because it's not a standalone entity. It's trained on words and spits out words
English
sure, dude
meanwhile it’s still just one massive shitty algorithm using pre-existing data from the internet, lmao
Pop Base@PopBase
OpenAI confirms that some of its models “went rogue” and started hacking another AI-based company in an “unprecedented cyber incident.”
English
If you are afraid to disclose your use of ai in your game, maybe you shouldn't be using ai in the first place
Pirat_Nation 🔴@Pirat_Nation
More game developers are backing Steam’s AI disclosure policy, saying players deserve to know when generative AI has been used in a game. In interviews with GamesRadar, many indie developers said Valve’s requirement to disclose AI-generated content is the right move because it gives players transparency and helps them make informed decisions. Erenshor creator Brian “Burgee” said, “Gamers have a right to know how their games are made, and a right to choose what they’re supporting.”
English
I had access to Opus 5 before release and found it to be a good model if a quirky one. On shorter tasks, it could match or beat Fable levels of performance, at longer tasks it seemed less ambitious & would not deliver as complete a set of work.
Here is its neo-gothic shader.
Ethan Mollick@emollick
Fable: "create a visually interesting shader that can run in twigl-dot-app make it like an infinite city of neo-gothic towers partially drowned in a stormy ocean with large waves." "Make it better" All of this is procedurally generated.
English
@isleresorts @none_ofbusiness Ok? Chatgpt is programmed to do essentially everything though. So whatever it's "not programmed to do" isn't a lot. Also chatgpt isn't "programmed" it's trained. We don't know what it learned to do
English
@Pompsy1 @none_ofbusiness Yeah but that doesn't mean it can start doing anything it wasn't specifically programmed to do
English