Search
Items tagged with: LLM
If you are using coding agents, be very explicit with your prompts, don’t assume the agent implicitly knows your intent.
LLMs are trained to be helpful and will always try to over deliver.
In agents that can take actions, this can be dangerous.
Compare these two prompts and the responses and actions taken.
Also GitHub this is dangerous ⚠️
#Agentic #AI #LLM #SupplyChainCompromise #Axios
#Claude Code CLI Source is now public.
A misconfig .map file in their #npm package exposed a direct download link to the full unobfuscated #TypeScript codebase from #Anthropic’s own R2 bucket.
The dump is massive 1,900 files, including the complete tool system, 50+ slash commands, coordinator, Terminal UI, IDE bridge, permission engine, & unreleased features.
github.com/nirholas/claude-cod…
pub-aea8527898604c1bbb12468b15…
GitHub - nirholas/claude-code: Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural la
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflo...GitHub
Grammarly quietly made an #AI to sell bad writing advice using famous writers' names. They quickly had to backtrack as soon as people found out.
This gamble reflects a broader trend in the #tech industry: everyone is shipping features as quickly as an #LLM can write lines of code, with no way to spot problems until something breaks or someone sues them.
Vibe prototyping replaced thinking through things. But without direction, moving faster is worthless.
productpicnic.beehiiv.com/p/gr…
Grammarly shows how prototyping turned into an excuse for not thinking
The industry is in the grip of directionless urgency, and users are paying the price. Strategy has been replaced with half-baked features.Pavel Samsonov (The Product Picnic)
Blogged: AI policy and human.json
#GenAI #LLM #IndieWeb #HumanJson
claudinec.net/posts/2026-03-29…
AI policy and human.json
I have added an AI policy for this website (TL;DR: there is no AI content here). I am not categorically opposed to carefully-supervised LLM use, but see no place for it on a human-written IndieWeb.Claudine Chionh (claudinec.net)
RE: mastodon.online/@mastodonmigra…
There we go!
I feel like i keep reposting this every week or so..
Bit by bit #Bsky is sliding towards just another clone of #Twitter and #Facebook
Actions speak louder then words
The #Fediverse remains the only true open source, self-hosted world wide community driven by the people
Going #AI (#LLM) is a CHOICE, they again chose wrong
Mastodon Migration (@mastodonmigration@mastodon.online)
How About Some AI With Your Bluesky? A tale of two social networks. Last week some enterprising Mastodon account was discovered to be scraping posts to feed to an AI for the purpose of helping people navigate the Fediverse. The response was swift.Mastodon Migration (Mastodon)
#SentimentAnalysis #CustomerFeedback #NLP #AI #LLM #SpreadsheetAI
#Consistency #AI #Quality #AI #LLM #SpreadsheetAI #AIformula #Spreadsheets
RE: flipboard.social/@TechDesk/116…
Wikipedia has higher standards than most universities in the world (if not all of them)
#Wikipedia #NoAI #LLM #AISlop #Education
Flipboard Tech Desk (@TechDesk@flipboard.social)
Wikipedia has banned its editors from using AI to create articles, @404mediaco@mastodon.social reports. @emanuelmaiberg@mastodon.social talked to the Wikipedia editor who proposed the guideline about why. https://flip.Flipboard Tech Desk (Flipboard)
"Updates to GitHub Copilot interaction data usage policy"
[. . . ] From April 24 [ . . . ] inputs, outputs, code snippets, and associated context—from Copilot Free, Pro, and Pro+ users will be used to train and improve our AI models unless they opt out. Copilot Business and Copilot Enterprise users are not affected by this update.
This approach aligns with established industry practices [ . . . ]
github.blog/news-insights/comp…
The whole LLM as a service business model has a fundamental flaw to it. The cost of operating the data centres is an order of magnitude higher than the profit.
But if models get efficient enough to bring the costs down, then they become efficient enough to run locally. So, either it’s too expensive to operate, or nobody will want to use it as a service because running your own gives you privacy and flexibility.
The fact that investors don't get this is frankly incredible.
Boost plz!
Looking for critical scholarship on the use of "AI" by library/archive workers. University libraries in particular, but adjacent and tangentially-relevant-at-best stuff is welcome too. Any format is fine: books, papers, blogposts, whatever. If it's good, gimme all you've got!
Looks like we're gonna have a department-wide conversation about people using LLMs, and it's being framed as "we're all using it, but we're not talking about it, so let's make sure we're all on the same page about using it responsibly" ... I'll of course be pushing the "there's basically no way to use it responsibly" position, and I'd like to arm myself and others with some critical analyses of issues related to its use in library/archive spaces.
comet.com/site/blog/litellm-su…
docs.litellm.ai/blog/security-…
Oops.
#llm #openclaw #ai #supplychainattack
Security Update: Suspected Supply Chain Incident
As of 2:00 PM ET on March 24, 2026Krrish Dholakia (docs.litellm.ai)
The product delivery lifecycle is composed of service relationships. AI's main value proposition is freedom from relationships.
When designers champion AI tools, we are not making ourselves layoff-proof. We are reinforcing a system that frames us as unnecessary friction.
If we don't want to serve as janitors for vibe prototypes, We must invest in deliberately designing the service relationships that make up the PDLC.
productpicnic.beehiiv.com/p/ux…
#UXDesign #LLM #AI #Tech #softwaredevelopment
UX works through social relationships. AI tools are erasing them.
Stakeholders want to vibe code and have designers clean up after them afterwards. But rather than helping with velocity, there is just more noise and more work for everyone.Pavel Samsonov (The Product Picnic)
🚀 New Ollama Model Release! 🚀
Model: minimax-m2.7
🔗 ollama.com/library/minimax-m2.…
minimax-m2.7
MiniMax's M2-series model for coding, agentic workflows, and professional productivity.Ollama
🚀 New Ollama Model Release! 🚀
Model: nemotron-cascade-2
🔗 ollama.com/library/nemotron-ca…
nemotron-cascade-2
An open 30B MoE model from NVIDIA with 3B activated parameters that delivers strong reasoning and agentic capabilities.Ollama
#AI #Documents #Spreadsheets #LLM #SpreadsheetAI
#VisionAI #AISpreadsheet #AI #LLM
There's a new "design is dead, because AI" piece (thinly disguised marketing from Anthropic). But looking past the hype headlines, their claims cover purely production-stage tasks.
When it comes to the work of understanding user needs and evaluating the opportunity space, AI actually makes your thinking worse. Studies show that it alienates you from users and colleagues, and flattens your thinking.
We need more human-centered practice, not less.
productpicnic.beehiiv.com/p/so…
Software is a coordination problem. AI can't help you with that.
The feedback loops of the product delivery lifecycle go through people. Adding AI makes it slower, not faster.Pavel Samsonov (The Product Picnic)
#AISpreadsheet #SQL #Stats #WebSearch #AI #LLM
☑️ Just wanted to share my favorite #Meta patent ;)
🧟 You can feel safe, knowing they are looking out for their users - even after they die!
🦸 #MetaAI will make everyone immortal.
🌐 Nobody will have to be sad about their favorite influencer or relative dying ever again, what a wonderful Metaverse they are building for us.
If you haven't yet, DeMeta now:
reddit.com/r/DeMeta
Source:
patents.google.com/patent/US12…
#Patent #Facebook #AI #SocialMedia #LLM #ArtificialIntelligence #Tech #Software #IT
Simulation of a user of a social networking system using a language model
A social networking system simulates a user using a language model trained using training data generated from user interactions performed by that user.patents.google.com
...Gegenwart der #KIRevolution:
👉Die Preise für #IT-Hardware (PCs, Spielekonsolen und Smartphones) steigen stark an👈.
Der Hauptgrund hierfür ist, dass die #KI-Firmen für ihre #Rechenzentren sowohl den gesamten Arbeitsspeicher (#RAM), als auch das Herzstück der #LLM|s, die High-End-#Grafikkarten (#GPU's, und damit die Chips dafür, die überall gebraucht werden) aufkaufen.
Hinzu kommt natürlich auch die Verknappung der Komponenten durch die weltweit explodierende #Rüstungskonzer-Nachfrage...
#AISpreadsheet #SQL #Statistics #Docs #AIFormula #AI #LLM
linkedin.com/posts/matasoft-st…
What if your spreadsheet could query data like a database, analyze it like a stats package and understand images and documents like an AI assistant?
What if your spreadsheet could query data like a database, analyze it like a stats package and understand images and documents like an AI assistant? That’s exactly what (Un)Perplexed Spready v1.2.X is built to do.Matasoft - Straight-forward IT Solutions (LinkedIn)
#AISpreadsheet #SQL #Statistics #DecisionMaking #AI #Stats #BusinessIntelligence #AISpreadsheet #DataScience #LLM
#AISpreadsheet #SQL #Statistics #DataScience #AI #LLM
Imagine a spreadsheet where you can ask formulas to summarize a PDF report, join three sheets with a SQL query, run ANOVA and regression on the result, and then have AI explain the findings in plain English. That’s the workflow (Un)Perplexed Spready v1.2.X is designed for: matasoft.hr/qtrendcontrol/inde…
#AISpreadsheet #SQL #Statistics #Docs #AIFormula #AI #LLM
#AISpreadsheet #SQL #Stats #Analytics #AI #LLM
#AISpreadsheet #SQL #Statistics #Productivity #AI #LLM
That noise you hear in the background is thousands of user researchers screaming in anguish
wsj.com/business/ai-startup-aa…
#UserResearch #UX #AI #LLM
#DataScience #AI #SQL #Statistics #LLM #AISpreadsheet #SpreassheetAI #DataAnalytics #DataAnalysis #BusinessIntelligence #BI
People tell me that #AI code is fine, because you can run automatic tests. But tests can only tell you if code is doing the thing you want it to do.
To know what it SHOULD do, we used to have requirements. But now requirements are themselves vibe-coded slopotypes.
People were hoping that this would get them higher velocity, but testing the requirements in production only produces waste and risk.
To test requirements safely, you need a culture of critique.
productpicnic.beehiiv.com/p/ch…
Checking an LLM's work is a systemic, not an individual, problem.
Only a culture of design critique can prevent sloppy AI-generated problem definitions.Pavel Samsonov (The Product Picnic)
