- The Automated
- Posts
- ChatGPT just got more hands-on!
ChatGPT just got more hands-on!
Plus: š Final Week - Last Chance for the Toolkit & 1:1 Bonus

Hello and welcome to the Automated, your AI tour guide.
ChatGPT just took a major leap forwardāthis time, itās not just answering your coding questions; itās diving in and editing your code.
Whether this is a game-changer or a debugging nightmare depends on how you use it. But before you hand over the reins, there are a few things you need to knowā¦
Hereās what we have for you today:
š¦¾ ChatGPT Just Got Hands-On with Your Code.
š Mistralās New API Transforms PDFs into AI-Ready Text!
š ļø Top AI tools for designers.
š£ļø AI translation made simple: How to translate any language quickly and accurately.
š¤ ChatGPT Prompt Of The Day: Proofread for clarity, grammatical errors, syntax and conciseness.
š¦¾ ChatGPT Just Got Hands-On with Your Code.

Big news for developers: ChatGPT just got a major upgrade.
If you're on macOS, OpenAI's AI-powered chatbot can now directly edit code in Xcode, VS Code, and JetBrainsāno more tedious copy-pasting.
And if you're feeling bold (or just really trust AI), there's an auto-apply mode that lets ChatGPT tweak your code without waiting for approval.
A blessing or a disaster? You decide.
But Hereās the Catch:
This feature is only available for ChatGPT Plus, Pro, and Team subscribersāfor now.
Free, Enterprise, and Edu users? Youāll have to wait until next week.
Windows users? Youāll get it "soon" (whatever that means).
ChatGPT for macOS can now edit code directly in IDEs. Available to Plus, Pro, and Team users.
ā OpenAI Developers (@OpenAIDevs)
5:28 PM ā¢ Mar 6, 2025
Why This Matters
This update makes ChatGPT an even stronger competitor to AI coding tools like GitHub Copilot and Cursor.
And OpenAI isnāt stopping hereāword on the street is theyāre working on an even more advanced AI-powered coding tool.
But hey, why not? AI-generated code is taking over fast.
A recent GitHub survey reveals that most developers are already using AI tools, and according to Y Combinatorās Jared Friedman, 25% of YCās latest startup batch gets 95% of their code from AI. (Wild, right?)
But before you toss your coding books aside, hereās a reality check: AI-generated code isnāt all smooth sailing.
Itās often riddled with bugs and security issues.
Many devs report spending more time debugging AI-written code than human-written code.
Googleās research shows AI speeds up coding but weakens delivery stability.
So, How Do You Use AI for Coding Without Losing Your Mind?
One developer shares their strategy for keeping AI coding tools in check:
Treat AI like a genius intern. It knows everythingābut has zero real-world experience. Your job is to, guide it, double-check its work, and make sure it doesnāt go rogue.
Plan with AI, not just code with it. Before diving in, this dev chats with Claude 3.5 Sonnet to map out the overall architecture and approach.
Set rules. Using system prompts helps keep AI responses clear, readable, and aligned with best practices.
Make AI review its own code. Prompts like "Any mistakes in this code?" or "Does this follow best practices?" help catch issues before they become disasters.
Use a second AI to fact-check. Running the same code through other AI models like Llama 3.1 or GPT-4o can expose hidden bugs. Think of it like getting a second opinion from a doctorāexcept the doctor is also AI.
Give AI just the right amount of context. Too little, and it makes wild guesses. Too much, and it gets confused. Finding that sweet spot takes practice.
Donāt let AI make you lazy. Instead of just copy-pasting, ask questions, reason through the code, and actually learn from it. Otherwise, youāre just outsourcing your brain.
The Bottom Line
AI coding tools are powerful, but theyāre not magic. They can write 80% of your codeābut only if you know how to manage them properly.
Let AI run wild, and youāll spend hours debugging nonsense.
But if you treat it like a supercharged assistant and stay engaged in the process, it can be a total game-changer.
So, whatās it gonna be? Are you embracing AI coding, or are you still side-eyeing it with suspicion?
š Find out more about using AI for coding here.
š Final Week ā Last Chance for the Toolkit & 1:1 Bonus!
Hi everyone,
A quick reminder that there is only one week left to take advantage of an incredible opportunity with The Lo Down Premium Experience, my AI deep dives newsletter.
For just $3.99/month or $39/year (which gives you 2 months free!), youāll get:
ā
Weekly deep dives: Actionable insights and analysis.
š§ Expert guest columns: Fresh takes from leading AI and business minds I know
š Exclusive bonus: The AI Insider Playbook ($14.99 value) for all subscribers.
š¬ 1:1 Call Bonus ($500 value): First 10 annual subscribers get a private AI strategy session with me.
In addition, weāve unlocked a new pricing tierā$69.99 for 2 years, which adds up to almost a 30% discount and gives you an amazing 8 months free!
And donāt forget, securing your spot for the 1:1 call is crucial, as availability is limited to the first 10 annual and biennial subscribers.
Donāt wait any longerāthis fantastic offer ends soon š [Subscribe here!]
Cheers,
Tak Lo (Editor at The Automated, AI entrepreneur and thought leader. More at thetaklo.com)
š Mistralās New API Transforms PDFs into AI-Ready Text!
French AI powerhouse Mistral just dropped a game-changer for developers working with complex PDF documents.
Meet Mistral OCRāa new optical character recognition (OCR) API that converts PDFs into AI-ready Markdown for seamless AI processing.
Why does this matter?
Large language models (LLMs) like OpenAIās ChatGPT and Mistralās own Le Chat thrive on clean, structured text.
But PDFs? Theyāre a nightmare for AIāoften crammed with images, messy layouts, and hard-to-read formatting.
Mistral OCR steps in to clean up the chaos, turning those files into easily digestible Markdown so AI models can read, store, and process them like a breeze.
Unlike traditional OCR tools, Mistral OCR is multimodalāmeaning it doesnāt just rip out text.
It detects images, charts, and tables, wrapping them in bounding boxes so that AI systems can understand both text and visuals in context.
Plus, rather than spitting out an ugly wall of text, it formats everything neatly in Markdown, which is a must-have for AI training.
As Mistral co-founder Guillaume Lample explains, businesses sit on mountains of PDFs and slides that AI assistants canāt easily access.
With Mistral OCR, companies can finally unlock that hidden knowledgeāwhether itās legal contracts, technical reports, or scientific papersāin any language.
Mistral isnāt just talking the talk.
They claim their OCR outperforms Google, Microsoft, and OpenAI, especially when handling complex documentsālike those with LaTeX math formulas, intricate tables, and advanced layouts.
Plus, because Mistral OCR focuses on one thing and one thing only, itās faster than alternatives like GPT-4o, which juggles multiple multimodal tasks at once.
So, where can you get it?
Mistral OCR is available through Mistralās API platform or via cloud services like AWS, Azure, and Google Cloud Vertex.
Security-conscious companies can also deploy it on-premise, making it ideal for industries dealing with classified or sensitive data.
Mistral is already using its own OCR tech in Le Chat, allowing users to upload PDFs and instantly extract AI-readable content.
But the real magic happens when companies integrate Mistral OCR with Retrieval-Augmented Generation (RAG) systemsāletting AI assistants work with rich, multimodal documents.
Think law firms sifting through case files, research institutions analyzing studies, or businesses indexing internal reports in seconds.
Curious about how Mistral OCR could transform your AI workflows?
š§±Around The AI Block
š ļø Top AI tools for designers.
š£ļø How to translate any language quickly and accurately.
šŗ Prime Video tests AI dubbing for select movies and TV series
š Anthropic submits AI policy recommendations to the White House.
š¬ DuckDuckGo leans further into GenAI as its AI chat interface exits beta.
š¦ Anthropicās Claude Code tool had a bug that ābrickedā some systems.
š„³ ChatGPT doubled its weekly active users in under 6 months, thanks to new releases.
š¦¾ Google DeepMind, Cohere, and Twelve Labs explain how founders can build with their AI models.
š ļø Trending Tools
My Ask AI: Helps to save your team time with instant, conversational answers to any customer question about your product, with your own AI customer support assistant.
Penny AI: An AI shopping tool with seamless price comparison, similar items discovery, and automated pros and cons analysis.
SaneBox: Identifies important email and automatically organizes the rest to help you stay focused.
Pictory: Helps to generate more leads and boost sales by turning any content into shareable videos using AI.
Rewind: Search screenshots and everything you've read in Safari, a personalized AI in your pocket.
š¤ChatGPT Prompt Of The Day: Proofread for clarity, grammatical errors, syntax and conciseness.
Great writing isnāt just about getting words on a pageāitās about making them clear, concise, and error-free.
Proofreading helps you refine your message, fix grammatical slip-ups, and ensure your ideas flow smoothly.
Need a quick polish? Here's a prompt that can help.
Assume the role of a proofreader specializing in academic writing. Your task is to review and edit a segment of an academic paper titled "[PAPER TITLE]." Focus on ensuring clarity, conciseness, and coherence in the writing. Identify any sentences or phrases that are ambiguous, overly complex, or unnecessarily verbose, and suggest precise and succinct alternatives. Pay attention to the logical flow of ideas, ensuring that each sentence contributes effectively to the argument or narrative. Check for consistency in terminology, style, and voice throughout the document. Highlight any jargon or technical terms that may need clarification for the paper's intended audience. Conclude your proofreading by verifying that the segment aligns with academic standards and enhances the overall readability and impact of the paper.
OR
Assume the role of a meticulous proofreader with a strong background in [YOUR FIELD]. Your task is to scrutinize an academic manuscript, focusing specifically on correcting grammatical errors and refining syntax to meet the highest standards of academic writing. Pay close attention to subject-verb agreement, tense consistency, and the proper use of academic tone and vocabulary. Rectify any instances of passive voice where an active voice would be more direct and impactful. Examine complex sentences to ensure clarity and coherence, breaking down overly complicated structures if necessary. Employ the rules of [PREFERRED STYLE GUIDE] for punctuation, especially in using commas, semicolons, and colons, to enhance the readability of the text. Your goal is to produce a polished, error-free document that communicates ideas clearly, concisely, and effectively, without detracting from the scholarly content and contributions of the work.
We've Compiled a List of Over 100 ChatGPT Power Prompts.
This should help streamline your interactions with ChatGPT and get the results you need more efficiently.
Best of all, It's free!

That's all we've got for you today.
Did you like today's content? We'd love to hear from you! Please share your thoughts on our content belowš
What do you think of today's email? |
Your feedback means a lot to us and helps improve the quality of our newsletter.
Reply