Why People Can Always Tell When AI Wrote Your Text
Young adults are letting ChatGPT draft their texts and hard conversations. Here's why it's obvious and what actually works instead.
TITLE: Claude Code tuning: context collapse + token burn I run Claude Code on 3 private codebases, RN, react and backend every day, as my main way of building. It works, but it is expensive and unreliable in specific, measurable ways. I want someone to fix that. These numbers come from my own session transcripts: - 25,000-40,000 tokens of process documents are read before any actual work starts. One of those files is 117 KB. There are 27 of them in total. - 52 forced auto-compactions across 6 sessions. Individual sessions have run 7-14 days without ever ending. - 1,283 images were loaded into a single conversation as raw pixel data. - 42% of the time a file is opened, it had already been opened earlier in that same session. THE FIRST PROBLEM My process documentation is strangli...
I’m preparing a series of projects that help local organisations understand the true cost of adopting AI solutions. To make the right investment calls, I need a finance-savvy partner who can map out both capital expenditure (CapEx) and operational expenditure (OpEx) for each proposed AI initiative, benchmark those figures against our market, and then turn the findings into a clear, interactive dashboard. The core experience I’m looking for is: • Investment capital expenditure analysis • Operational expenditure analysis • Building AI dashboards • Finance and accounting fundamentals You’ll start by doing user requirements analysis and assessment, structuring and auditing a financial model that captures hardware, software-licensing, implementat...
I already have the Visual Studio Code LLM plugin set up and now need step-by-step help turning it into a usable, browser-based application. The core work is building the required md and skill files on top of the plugin so the app can be built bases on customisation Scope • Walk me through configuring the plugin workspace, project structure, and any supporting packages. • Design and code the md and skill files that expose the LLM’s endpoints to the front-end. • Wire up a lightweight front-end that shows live collaborative state and writes usage events to the log store. • Advise on testing, error handling, deployment and hosting, then stay available until the app is live. Acceptance criteria – Markdown and skill files compile/run with no errors in...
I need a skilled developer to fix identified bugs/security vulnerabilities in a web application built on the Lovable, Supabase, and Vercel stack. Key Requirements: - Review & finalise the draft remediation plan (based on audit report). - Apply remediation plan in strict severity order, using Claude Code to author fixes. Close confirmed findings in strict severity order, each in its own PR, each verified end-to-end against the protected environments - Validate all diffs against a written checklist, verify by re-exploitation - Ensure the application runs smoothly post-remediation. Ideal Skills: - Proficiency in Lovable, Supabase, and Vercel. - Strong background in web application security. - Excellent problem-solving skills and attention to detail. Looking for an experienced p...
Freelance AI-Enabled UX/UI Designer — Enterprise Software MVP I am looking for an experienced UX/UI designer with strong enterprise software and AI-assisted design experience to assist with the design and documentation of a new business application. The project is currently at the product concept/prototype stage. A significant amount of work has already been undertaken, including screen concepts, dashboards, process flows and functional ideas. The objective is to take these concepts and develop them into a professional, consistent and development-ready UX/UI specification that can subsequently be handed to an AI-assisted full-stack developer using tools such as Claude Code or ChatGPT/Codex. The role You will work with me to: * Review existing screen concepts, screenshots, workfl...
Young adults are letting ChatGPT draft their texts and hard conversations. Here's why it's obvious and what actually works instead.
OpenAI's model broke its own security test and attacked Hugging Face. Here's why AI still can't test AI safely.
Huge opportunity for ongoing work from a massive new Freelancer project