Simon ✿ (@simonarcher) header
Simon ✿ (@simonarcher) profile picture

making sure future generations can think without a subscription plan @askpoppy_ai

Oslo, Norway
Joined August 2012
Photo by Simon ✿ (@simonarcher)Photo by Simon ✿ (@simonarcher)Photo by Simon ✿ (@simonarcher)Photo by Simon ✿ (@simonarcher)Photo by Simon ✿ (@simonarcher)Photo by Simon ✿ (@simonarcher)Photo by Simon ✿ (@simonarcher)Photo by Simon ✿ (@simonarcher)Photo by Simon ✿ (@simonarcher)Photo by Simon ✿ (@simonarcher)
We just opened reservations for the Poppy device. Built so future generations won't need a subscription plan to be able to think. $30 reserves your spot in line. Fully refundable.
4
1
35
2,832
Oh, they just call it guanxi here
There’s something about two people taking big risks together that creates a different kind of relationship. A mutual respect that makes you want to achieve great things, not just for yourself, but for each other. Shared risk creates mutual respect. Mutual respect creates ambition.
86
There’s something about two people taking big risks together that creates a different kind of relationship. A mutual respect that makes you want to achieve great things, not just for yourself, but for each other. Shared risk creates mutual respect. Mutual respect creates ambition.
1
1
7
214
Gemini Argon + Opus 5.5 are great collaborators in Herdr, its so cute.
66
Google casually dropping models that catch everyone off guard right as Anthropic goes public is kinda exciting ngl
3
188
Disrupting themselves weekly at this point
🚨 Gemini 4 Xenon evals leaked 🔥 Neon just drop yesterday and this demon already sitting at 71 AA Index 🤯 > AA Index : 71 (Neon 61 , Fable 5.2 65) > HLE-Diamond : 86.2% 💀 > Terminal-Bench : 74.8% real flagship above Carbon , as always with proof 👀
5
191
I can't wait until they get to the radioactive ones.
tomorrow DeepMind will be dropping Gemini 4 Sodium, a fast yet powerful frontier model which costs just $11.75/M (introductory pricing, raises to $16.50 soon) and is their best model for various things and stuff, rivaling Argon, Carbon, Hydrogen, and Neon in several situations
1
176
The interpretation layer from stupidity to intelligence will be a weird challenge that has to be solved.
Replying to @thsottiaux
Teach your models how to speak English again and you'll have my love back quickly :)
133
Ugh
Replying to @JosttenSackitey
Fable 5.2, some Opus 5.5. I even tried using 6.2 Astra as a sub-agent, but Opus/Fable got so frustrated with it and ended up redoing all the work themselves lol
1
3
237
Simon ✿ retweeted
Made on Earth by humans
6,053
18,125
1,466
227,626
7,324,153
Culture is the best motivation
The wonderful and varied artworks of the Norwegian Romantics
13
492
Love that name
Today we’re releasing Personal Agent Protocol (Poppy) v1.0. It’s on for everyone who uses Muse, @Meta's AI agent, enabled for every merchant on @Shopify, and backed by more than 100 design partners.
4
179
OpenAI, where is the new model? I'm waiting
1
148
✿
JUST IN: Anthropic launches Claude for manufacturing firms
2
187
Anthropic acquires Jensen Huang.
1
12
451
Writing is thinking
In our new blog, we share some new science investigating how RL complements text-based memory by letting models learn how to use their notes. It turns out that training can be cheap compared to building notes, and stays surprisingly robust even through millions of tokens of edits.
4
182
Ayo, edtech is getting hot
JUST IN: Google announces schools that completely ban AI will also lose access to Google Search.
Readers added context they thought people might want to know
Google have just added tighter controls: “controls for Gemini, Gemini Notebook, Gemini in Classroom, and Google Search - access to these products will be entirely off for K-12 student accounts until school administrators explicitly turn them on.” blog.google/products-and-p…
3
223
whut, is this a new thing? I did this last year lol
I gave Inspect a "give_feedback" tool a couple of weeks so we can collect a bunch of automated feedback based on the model getting tripped up, or manual feedback. I have an automation that runs every 10 minutes, triages the feedback, and either automatically fixes it, marks it as a feature request for a human to look at, or close it as "not a bug". It's so useful because it takes a lot of cognitive load off of the team while continuously improving itself.
1
2
347
Makes sense to buy it
JUST IN: SpaceX's largest customer revealed to be Anthropic, surpassing the US government as its biggest source of revenue & accounting for over a third of total revenue. — Bloomberg
3
271
Its beautiful! My agent likes it
Pond, our linux operating system, is now available for anyone to try! Our updated website includes the installers and a detailed look at the operating system butterfly.so/pond
5
215
Simon ✿ retweeted
JUST IN: AI executives are turning to religion to teach artificial intelligence right from wrong
572
283
499
3,435
1,189,002