Tootfinder

"Does adding ‘please’ and ‘thank you’ to your ChatGPT prompts really waste energy?"
#AI #ArtificialIntelligence #Energy

If you want to understand the importance of prompts in AI, it is interesing to read all Anthropic system prompts which are actually published (other companies don't). Simon Willison gave me the tip in his excellent blog. https://simonwillison.net/2026/Apr/18/opus-system-prompt/

Mittwoch: Warnung vor deutscher Superbehörde, gespeicherte KI-Prompts in Chrome
Risiko durch zentralisierte Kontrolle Google-Skills für Chrome Copilot statt ChatGPT in Norwegen Linux 7.0 mit Neuheiten Unitree-Roboter als Sprinter

As political pressure prompts exodus of Minnesota prosecutors, some defendants catch a break (Associated Press)
https://apnews.com/article/minnesota-prosecutors-office-defendant-free-142f82b163a8b0d646ba6079251d32be
http://www.memeorandum.com/260219/p88#a260219p88

This study's results suggest that LLM sycophancy amplifies publication bias (the "file drawer" problem)
The LLMs didn't like to tell the user that their hypothesis was not confirmed by the data
H/t @…

«Nie wieder denselben Prompt tippen — Google #Chrome speichert deine KI-#Workflow's als Skills:
Mit Google Skills kannst du deine wichtigsten Prompts als #Skills speichern und mit

Google's TIG says Gemini has been inundated by "commercially motivated" actors who are trying to clone it, including one campaign that prompted it 100K times (Kevin Collier/NBC News)
https://www.nbcnews.com/tech/security/google…

Automattic adds an AI assistant to WordPress.com, enabling users to perform site-wide layout and style modifications via natural language commands (Sarah Perez/TechCrunch)
https://techcrunch.com/2026/02/17/wordpress-com…

KI sollte Arbeit erleichtern. Stattdessen sitzen viele zwischen ChatGPT, Claude & Co. und merken abends: Gehirn weichgekocht. „Brain Fry“ nennt das eine Studie von BCG & University of California Riverside. Schuld ist nicht die KI – sondern unser Tool-Zapping und Optimierungsdrang. Weniger Prompts, mehr Pause.
KI grillt dein Gehirn - Jens Polomski
#KI

#ClaudeCode system prompts
https://platform.claude.com/docs/en/release-notes/system-prompts

A few months ago, I setup Pika Backup and Nextcloud online account on my mini PC running bazzite-deck-gnome. Both applications store their secrets in a gnome-keyring which didn't get unlocked because there was no password prompt on "Switch to desktop", so I was pestered every time with two different prompts for unlocking the same keyring, none of which I could just dismiss in favor of the other.
Today, I finally looked into if and how I could rebase to proper GNOME GDM lo…

Nutzer können KI-Anfragen bald in Chrome speichern für schnelle erneute Nutzung
Chrome erlaubt künftig das Speichern oft genutzter KI-Prompts im Browser, um Nutzern das erneute Eintippen der Befehle zu ersparen. Google nennt dies „Skills“.

Indian AI lab Lossfunk's prompting method lets LLMs generate Tulu language text without prior training; the method may expand to other low-resource languages (The Economic Times)
https://economi…

🐮 Feral horses and cattle create more resilient nature, rewilding study reveals
#conservation

Republican congressman's anti-Muslim remark prompts calls for his resignation (Ben Kamisar/NBC News)
https://www.nbcnews.com/politics/congress/randy-fine-anti-muslim-post-on-x-dogs-calls-for-resignation-rcna259270
http://www.memeorandum.com/260217/p21#a260217p21

<broken record>Please post your code and prompts for your vibecoded app for your twitter clone for your 20 friends or whatever you are humble bragging on the internet you and, i say this in the most respectful way, you "c*wards" :-)</broken record>

Automattic adds an AI assistant to WordPress.com, enabling users to perform site-wide layout and style modifications via natural language commands. (Sarah Perez/TechCrunch)
https://techcrunch.com/2026/02/17/wordpress-co…

Patrick Drahi's new ~15% stake in Channel 13 prompts questions about media independence in Israel; Drahi owns the outlet i24, which supports Benjamin Netanyahu (Financial Times)
https://www.ft.com/content/55055c78-a50a-4b87-9667-8081c8944c15

Too late, he's doubled down: “That staffer should be fired immediately, and the President should publicly condemn this action,” Burns continued. “This kind of insensitive and racist communication does not reflect the heart, values, or leadership of the President of the United States, nor does it represent the America we are striving to build.”
Trump-allied pastor: Staffer who posted Obamas video 'should be fired immediately'
https://thehill.com/blogs/in-the-know/5727010-racist-video-prompts-outrage/

AI / LLM sycophancy as a liability:
'A 2025 study by Fanous et al. tested GPT-4o, Claude Sonnet, and Gemini 1.5 Pro across math and medical domains. The results: these systems changed their answers nearly 60% of the time when challenged by users'
https://

FBI Subpoenaed X to Get Grok Prompts Used to Create Nonconsensual Porn https://www.404media.co/fbi-subpoenaed-x-to-get-grok-prompts-used-to-create-nonconsensual-porn/

More on why trusting AI to decide literally anything is a horrible idea.
https://www.randalolson.com/2026/02/07/the-are-you-sure-problem-why-your-ai-keeps-changing-its-mind/

Private funding of public art prompts questions about intent | Culture | sfexaminer.com
https://www.sfexaminer.com/culture/private-funding-of-public-art-prompts-questions-about-intent/article_89b…

Private funding of public art prompts questions about intent | Culture | sfexaminer.com
https://www.sfexaminer.com/culture/private-funding-of-public-art-prompts-questions-about-intent/article_89b…

@… i was more looking for "tools" not for "prompts/skills" - but sure, i take that to my list 😊

🍮 Wissen zum Nachtisch: 🍨
Immer mehr Menschen sehen sich im beruflichen Umfeld genötigt, mit generativer #KI zu arbeiten.
Besonders Großunternehmen „überrollen“ damit ihre #Mitarbeiter. Es wird eine Art

#OpenAI is looking for money. Subscription fees and selling ads is not enough.
Now they try to get a cut from the money you make thanks to #AI.
This also implies that any promises ever made not to look into your prompts and answers may be moot.

Court records: the FBI subpoenaed X for details about Grok prompts a man allegedly used to create 200 sexual deepfakes of a woman he knew in real life (Jason Koebler/404 Media)
https://www.404media.co/fbi-subpoenaed-x-to-get-grok-prompts-used-…

Looks like the vibe coders at Microsoft forgot to add "don't introduce command injection vulnerabilities" to their prompts?
https://www.cve.org/CVERecord?id=CVE-2026-20841

pkg upgrade -Fqy
For a single run, with conflicts, there may be multiple prompts for one more solver iteration.
I think this is negligible, not worth reporting as an issue. @… do you agree?
Thanks
#FreeBSD

Canva unveils Canva AI 2.0, which can generate editable layered designs from conversational prompts using Canva's foundation model built specifically for design (Olivia Poh/Bloomberg)
https://www.bloomberg.com/news/features/20

YouTube rolls out an AI-powered playlist generator that lets Premium and Music Premium users on iOS and Android use text prompts to create playlists (Ivan Mehta/TechCrunch)
https://techcrunch.com/2026/02/10/youtube-rolls-out-an-ai-playlist-gene…

'A recent post by Marcelo Emmerich proposes replacing package managers with a “prompt registry.” Instead of publishing code, library authors would publish AI prompts.'
Also, the Wikipedia article about sniffing your own farts shows this guy's portrait.
(quote from this post: https://nesbitt.io/2026/01/30/will-ai-make-package-managers-redundant.html)

I've been hacking on Beowulf again, today; and I'm reminded how enchanted I am with the Beowulf logo, which sort-of designed itself.
I take no credit for it, really.
#Lisp 1.5
https://git.journeyman.cc/simon/beowulf

Tim Cook Attends White House Screening of 'Melania' Movie, Prompts Calls for Apple Boycott
https://gizmodo.com/tim-cook-attends-white-house-screening-of-melania-movie-prompts-calls-for-apple-boycott-2000714041

Chase is now requiring five factor authentication. I log in with password on the website, which prompts me to open the mobile app, which requires a thumbprint. Then the app also demands a PIN. That's three separate challenges plus two persistent devices. Every single time I open the website, even in the same browser session.

This is no joke -- today's Metacurity is packed with the most crucial infosec developments you should know, including
--N. Korean hackers were behind malicious versions of Axios,
--Threat actors compromised Cisco with stolen creds from Trivy attack,
--Anthropic leaked part of Claude's source code,
--Apple alters its upgrade-or-stay vulnerable method,
--Recruiting startup Mecor hit by incident linked to LiteLLM attack,
--Vim and GNU Emacs text editor…

It's probably a lost cause and also too short notice, but I'm still thinking about #AlgoApril, which I proposed back in 2022, but it's an incomplete list of prompts and sadly never took off...
https://github.com/alg…

Great talk on using system prompts to improve LLM responses and reduce token usage. https://youtu.be/PlYAj9hK6BI?si=FDqo6IdHyfs9ZxEY

Google adds new features to its video editing app Vids, including directing and customizing avatars through text prompts and Veo 3.1 support (Ivan Mehta/TechCrunch)
https://techcrunch.com/2026/04/02/google-now-lets-you-direct-avatars-thro…

🍮 Wissen zum Nachtisch: 🍨
Immer mehr Menschen sehen sich im beruflichen Umfeld genötigt, mit generativer #KI zu arbeiten.
Besonders Großunternehmen „überrollen“ damit ihre #Mitarbeiter. Es wird eine Art

Clawdbot/Moltbot and the many security issues it introduces:

"This is not speculative. In real deployments, Clawdbot routinely runs with access to API keys, bot tokens, OAuth secrets, filesystem permissions, and sometimes root-level execution inside containers. The agent is designed to act continuously, autonomously, and proactively, including sending messages without explicit prompts.

This architecture is powerful, but it collapses several trust boundaries into a single …

RE: https://infosec.exchange/@haveibeenpwned/116322497293687889
Wow email addresses and prompts. So bad.

Why Pass@k Optimization Can Degrade Pass@1: Prompt Interference in LLM Post-training
Anas Barakat, Souradip Chakraborty, Khushbu Pahwa, Amrit Singh Bedi
https://arxiv.org/abs/2602.21189 https://arxiv.org/pdf/2602.21189 https://arxiv.org/html/2602.21189
arXiv:2602.21189v1 Announce Type: new
Abstract: Pass@k is a widely used performance metric for verifiable large language model tasks, including mathematical reasoning, code generation, and short-answer reasoning. It defines success if any of $k$ independently sampled solutions passes a verifier. This multi-sample inference metric has motivated inference-aware fine-tuning methods that directly optimize pass@$k$. However, prior work reports a recurring trade-off: pass@k improves while pass@1 degrades under such methods. This trade-off is practically important because pass@1 often remains a hard operational constraint due to latency and cost budgets, imperfect verifier coverage, and the need for a reliable single-shot fallback. We study the origin of this trade-off and provide a theoretical characterization of when pass@k policy optimization can reduce pass@1 through gradient conflict induced by prompt interference. We show that pass@$k$ policy gradients can conflict with pass@1 gradients because pass@$k$ optimization implicitly reweights prompts toward low-success prompts; when these prompts are what we term negatively interfering, their upweighting can rotate the pass@k update direction away from the pass@1 direction. We illustrate our theoretical findings with large language model experiments on verifiable mathematical reasoning tasks.
toXiv_bot_toot

„Vibe Coding XR“: Gemini generiert VR-Apps in unter einer Minute
Google stellt mit Vibe Coding XR einen KI-Workflow vor, mit dem sich aus Prompts VR-Apps für Android XR direkt im Browser generieren und testen lassen.

You don’t need to sign up to NBC to listen to articles, even if you use Firefox on Android.
I updated this post with the steps shown in the video (since Firefox has to rely on Android, versus its swell UI on Windows):
https://adrianroselli.com/2026/03/your-b…

YouTube rolls out an AI-powered playlist generator that lets Premium and Music Premium users on iOS and Android use text prompts to create playlists (Ivan Mehta/TechCrunch)
https://techcrunch.com/2026/02/10/youtube-rolls-out-an-ai-playlist-gene…

Turning off the computer without giving my agents a few prompts so they work the whole night.... Feels so bad...... 😅🙈 #ai #coding
Or should I be worried about, having that feeling in the first place, Damn, both doesn't feel healthy 😂🤔

Google launches Skills, repeatable AI prompts users can run in Chrome with a keyboard shortcut; users can set up their own Skills or choose from 50 presets (Reece Rogers/Wired)
https://www.wired.com/story/how-to-use-google-chrome-ai-powered-skills/

🍮 Wissen zum Nachtisch: 🍨
Immer mehr Menschen sehen sich im beruflichen Umfeld genötigt, mit generativer #KI zu arbeiten.
Besonders Großunternehmen „überrollen“ damit ihre #Mitarbeiter. Es wird eine Art

Sprache spielt auch beim Prompten eine Rolle.
Eine Studie von @… und @… zeigt, dass die Stärke von Geschlechterstereotypen in KI-Bildern je nach Sprache variiert.
Nicht überraschend: Prompts mit generischem Maskulinum erzeu…

With so many tokens that came before the current token, and so many possibilities that come after, it’s the job of the harness, system prompts, post-training, et al in concert with the human to weave that into something useful. Something valuable.
The LLM is the Temporal Loom.
And we’re The TVA.

I know many on here do not like Reddit. I mostly am on it for a few forums that do not exist or are dead on here.
And some light reading. I love to browse the Writing Prompts.
Go down to the post beginning with
The scream was not a sound Alton had ever heard a human being make.

Proc3D: Procedural 3D Generation and Parametric Editing of 3D Shapes with Large Language Models
Fadlullah Raji, Stefano Petrangeli, Matheus Gadelha, Yu Shen, Uttaran Bhattacharya, Gang Wu
https://arxiv.org/abs/2601.12234 https://arxiv.org/pdf/2601.12234 https://arxiv.org/html/2601.12234
arXiv:2601.12234v1 Announce Type: new
Abstract: Generating 3D models has traditionally been a complex task requiring specialized expertise. While recent advances in generative AI have sought to automate this process, existing methods produce non-editable representation, such as meshes or point clouds, limiting their adaptability for iterative design. In this paper, we introduce Proc3D, a system designed to generate editable 3D models while enabling real-time modifications. At its core, Proc3D introduces procedural compact graph (PCG), a graph representation of 3D models, that encodes the algorithmic rules and structures necessary for generating the model. This representation exposes key parameters, allowing intuitive manual adjustments via sliders and checkboxes, as well as real-time, automated modifications through natural language prompts using Large Language Models (LLMs). We demonstrate Proc3D's capabilities using two generative approaches: GPT-4o with in-context learning (ICL) and a fine-tuned LLAMA-3 model. Experimental results show that Proc3D outperforms existing methods in editing efficiency, achieving more than 400x speedup over conventional approaches that require full regeneration for each modification. Additionally, Proc3D improves ULIP scores by 28%, a metric that evaluates the alignment between generated 3D models and text prompts. By enabling text-aligned 3D model generation along with precise, real-time parametric edits, Proc3D facilitates highly accurate text-based image editing applications.
toXiv_bot_toot

Federal agent opens fire during immigration operation in Willowbrook; man taken into custody (ABC7)
https://abc7.com/post/shooting-involving-federal-agent-willowbrook-prompts-massive-response-authorities/18443548/
http://www.memeorandum.com/260121/p156#a260121p156

AI is just wonderful. A couple of prompts, and somethings that would have taken me ages before is done. Like this little macro that sends my model dierctly from #FreeCAD to OrcaSlicer.
Now it's just a click, and straight to slicing and printing.

Adobe updates Acrobat to add new AI features, including presentation generation, file editing, and creating podcast summaries from docs stored in Adobe Spaces (Ivan Mehta/TechCrunch)
https://techcrunch.com/2026/01/21/adobe-acr…

Spotify expands Prompted Playlists, its AI feature for creating playlists, to include podcasts, for Premium users in the US, Canada, the UK, and other countries (Ivan Mehta/TechCrunch)
https://techcrunch.com/2026/04/07/spotifys-prompted-pl…

Justice Thomas Bemoans Incivility as Security Prompts Cancellation of In-Person Speech (Ann E. Marimow/New York Times)
https://www.nytimes.com/2026/02/27/us/politics/supreme-court-thomas-security.html?unlocked_article_code=1.PVA.6dFp.5bx6M9PBLcBu&smid=url-share
http://www.memeorandum.com/260227/p79#a260227p79

Sources: WBD staff are bracing for widespread layoffs as David Ellison wants $6B in "synergies"; the internal mood is grim as staff preferred Netflix's proposal (New York Times)
https://www.nytimes.com/2026/02/28/business/warner-bros-paramount-re…

Luma AI debuts Uni-1, an image model that combines image understanding and generation in a single architecture, topping Nano Banana 2 on logic-based benchmarks (Matthias Bastian/The Decoder)
https://the-decoder.com/luma-ais-new-u

'Wotsit Street' scheme backlash prompts review
I dunno about Wotsits, to me it just looks like the road painter was having a really bad case of the hiccups.
#Birkenhead

Spotify expands Prompted Playlists, its AI feature for creating playlists, to include podcasts, for Premium users in the US, Canada, the UK, and other countries (Ivan Mehta/TechCrunch)
https://techcrunch.com/2026/04/07/spotifys-prompted-pl…

🔄 CacheAligner: Stabilizes prefixes so provider caching actually works - up to 10x savings by fixing dynamic content like dates that break cache hits
📊 RollingWindow: Manages context within limits without breaking tool calls - drops oldest tool outputs first, never breaks system prompts or recent turns
🔧 Easy setup options: Proxy server works with any #OpenAI-compatible client with…

The ICE killing of Alex Pretti, the second in Minneapolis, prompts quick legal action (Chris Geidner/Law Dork)
https://www.lawdork.com/p/the-ice-killing-of-alex-pretti-the
http://www.memeorandum.com/260125/p23#a260125p23

Profound, which tracks how AI models recommend brands across millions of prompts, raised a $96M Series C at a $1B valuation, bringing its total funding to $155M (Lily Mae Lazarus/Fortune)
https://fortune.com/2026/02/24/exclusi

Adobe updates Firefly to add Quick Cut, a new AI tool to edit footage and B-roll to create a first draft of the final video based on user prompts (Ivan Mehta/TechCrunch)
https://techcrunch.com/2026/02/25/adobe-fireflys-vi…

Sources: Amazon's "sponsored prompts" for its Rufus AI assistant are yielding significantly lower traffic than traditional ads, but are more cost-effective (Catherine Perloff/The Information)
https://www.theinformation.com/articles/amazons-ai-chat-a…

Shares of major gaming companies plunge after Google released Project Genie, which lets users create interactive worlds; Unity is down ~21% and Take-Two ~10% (Zaheer Kachwala/Reuters)
https://www.reuters.com/business/videogame

Profound, which tracks how AI models recommend brands across millions of prompts, raised a $96M Series C at a $1B valuation, bringing its total funding to $155M (Lily Mae Lazarus/Fortune)
https://fortune.com/2026/02/24/exclusi

Adobe updates Firefly to add Quick Cut, a new AI tool to edit footage and B-roll to create a first draft of the final video based on user prompts (Ivan Mehta/TechCrunch)
https://techcrunch.com/2026/02/25/adobe-fireflys-vi…

ADL study of Grok, ChatGPT, Llama, Claude, Gemini, and DeepSeek: Grok performed worst at identifying and countering antisemitic content, while Claude was best (Mia Sato/The Verge)
https://www.theverge.com/news/868925/adl-ai-…

OpenAI releases a set of prompts designed to be used with its open-weight safety model gpt-oss-safeguard that lets developers make their apps safer for teens (Amanda Silberling/TechCrunch)
https://techcrunch.com/2026/03/24/openai-adds…

Google rolls out Personal Intelligence in AI Mode, providing access to users' Gmail and Google Photos for more tailored responses, for US Pro and Ultra subs (Aisha Malik/TechCrunch)
https://techcrunch.com/2026/01/22/goog