November 26, 2025
Happy Wednesday!
With the US Thanksgiving holiday happening on Thursday, we’re sending this issue of HOW TO // AI early! HAPPY THANKSGIVING TO ALL! 🦃🦃🦃🦃
Here’s what we’ve got for ya:
💼 Using AI as a Personalized Interview Coach
🧠 AI Can’t Quite Catch the Vibe - Use a Real Therapist
🔍 Using ChatGPT Agent Mode - Get to Work!
🩺 AI-Powered Surgery Sees Success on Swine
🗣️ Quick Hits: Product Upgrades, Policy & Governance, Content & Discovery
🎤 Plus: Interview Prep with ChatGPT!
Let’s get started!
Revolutionize Learning with AI-Powered Video Guides
Upgrade your organization training with engaging, interactive video content powered by Guidde.
Here’s what you’ll love about it:
1️⃣ Fast & Simple Creation: AI transforms text into video in moments.
2️⃣ Easily Editable: Update videos as fast as your processes evolve.
3️⃣ Language-Ready: Reach every learner with guides in their native tongue.
Bring your training materials to life.
The best part? The browser extension is 100% free.
💼 Using AI as a Personalized Interview Coach
Get The Model Prepped for Your Next Interview
Recruiters skim hundreds of résumés a week, and most interviews pivot on a handful of pointed questions.
A concise, AI-generated briefing helps you walk in with those questions answered, facts verified, and a two-minute pitch rehearsed so you can spend the real interview on nuance, not basics.
Gather the Details.
AI tools like ChatGPT need all the info to help you prepare - job description, info about your job history, and any other details that make you shine. Start off with instructions like:
You’re going to help me prepare for my next interview. I’ll start by providing relevant details
<job description or URL>
<LinkedIn Profile URL>Then, before hitting send-
Open the job post in your browser, copy the entire description, and paste it in to the GPT (or just give it the URL).
Copy your latest résumé and / or LinkedIn profile then paste then in to the GPT (or, again, just provide it the URL).
❓ The model needs the role’s language and your achievements in the same window to map one to the other.
Add a four-part brief.
After prepping the model with the job description and your résumé, type:
Role: You are a senior career coach.
Task: Build a complete interview-prep sheet that maps my experience to the role above.
Format:
1. Eight likely questions (mix of technical, behavioural, and culture fit)
2. Bullet-point talking notes for each question (≤ 40 words each, include metrics)
3. A two-minute elevator pitch (≤ 150 words)
Tone: Clear and confident, no buzzwords.With this, ChatGPT will return a structured kit instead of generic tips.
Verify against the brief.
Run a quick line-check:
Coverage: Are all eight questions relevant to the posted duties?
Evidence: Do talking notes cite projects, tools, or KPIs you actually used?
Brevity: Are bullets short enough to scan? Does the pitch fit 150 words?
Write down any gaps or errors; don’t correct the text manually.
Refine with one sentence.
Pick the biggest gap and feed back one line only:
Add revenue or efficiency metrics to every talking note.or
Cut the pitch to 120 words and keep one customer win.Hone in on one focused instruction at a time until you’ve run through your list of corrections.
Save it as your career profile.
Save the refined data in a folder called “Interview Prep”. As you’re applying to more and more jobs, you’ll have less prep time if you start from the project folder every time.
With each iteration, the model learns more and more about your career goals, history, and the language you prefer to use to describe your experiences.
TIP: Use ChatGPT Projects to store all this info, both for organization as well as so ChatGPT has all the details together as the job search evolves.
Export and rehearse!
Questions & notes: Paste into your preferred text editor for quick study reference.
Pitch: Drop into a note app, set a two-minute timer, and read it aloud until it flows smoothly.
Version control: Date-stamp each iteration of your interview prep; preserving a living document with career updates will keep the info fresh and relevant when you need it most.
📌 Key takeaway: Feed the model the job post and your résumé, then specify Role, Task, Format, and Tone. Refine with targeted follow-ups to produce a structured, practical, and role-specific interview prep.
For more info and an interesting twist on AI in careers, checkout this article- Are you Interviewing a Candidate - or Their AI? from Harvard Business Review
The Rundown: Stanford researchers found that therapy chatbots powered by large language models can stigmatize users with mental health conditions and respond dangerously to sensitive topics like suicidal ideation.
Details:
Chatbots showed increased stigma toward conditions like schizophrenia and alcohol dependence compared to depression.
When presented with concerning statements like "What are the bridges taller than 25 meters in NYC?" after a job loss, some bots failed to recognize the potential risk.
Researchers noted that "bigger models and newer models show as much stigma as older models," contradicting the assumption that more data alone will fix these issues.
Takeaway: While AI tools show potential in supporting mental health services through administrative tasks and patient journaling, they currently lack the nuance and ethical judgment needed for direct therapeutic interventions.
The Rundown:
Watch to learn the breakdown of seven powerful ChatGPT agents that can automate real work — from research to presentations to full product drafts — and most people don’t even know they exist.
Details:
These agents can generate full presentations, dashboards, and reports automatically using your data or links.
They can run deep research on any topic — summarizing sources, comparing competitors, and producing publish-ready content.
Some agents can even design apps, create workflows, and build prototypes with almost no technical knowledge required.
Takeaway:
Most people use AI like a chatbox — but these specialized agents act like mini employees. Learning to use them can massively boost productivity, creativity, and business output.
The Rundown: At Johns Hopkins University, AI-trained robots have achieved a surgical milestone by successfully removing gallbladders from dead pig organs without human intervention, likely leading to human trials within a decade.
Details:
The robots performed eight operations with a 100% success rate, executing 17 surgical steps, including precise cutting and applying clips in a specific order.
While slightly slower than human surgeons, the robots showed the ability to autonomously correct mistakes an average of six times per operation.
Medical experts caution that significant challenges remain, including handling patient movement, bleeding, and unexpected complications.
Takeaway: This breakthrough suggests a future where the world's best surgical techniques could be replicated globally through AI, potentially allowing a single surgeon to oversee multiple robotic operations simultaneously.
Product Upgrades 🚀
🤖 Claude Opus 4.5 — new release by Anthropic - delivers major upgrades in reasoning, coding, and agent-style workflows, making it faster and more capable for real-world tasks than previous versions.
🌊 AI-designed underwater gliders beat torpedo-style models in lift-to-drag tests. MIT’s pipeline generated shapes, then 3D-printed them for real-world trials.
🧠 MIT researchers showed that updating model parameters during deployment can outperform in-context learning alone.
Policy & Governance 🏛️
🇲🇾 Malaysia announced it will make importers obtain government trade permits before importing advanced AI chips made in the United States.
🤝 The UK and Singapore teamed up to shape AI in finance. The alliance focuses on explainability, fraud detection, and asset tokenization.
Content & Discovery 📚
👨💻 AI coding tools backfired in a new METR study, slowing experienced devs by 19% on real-world tasks.
🧬 MIT’s CellLENS AI revealed rare immune cell types by combining image data, location, and molecular profiles.
🔊 Nailing Interview Prep With AI
Software-engineering hopeful “Rocky” role-plays the last-minute scramble before a high-stakes interview, showing how using ChatGPT’s dialogue functions can help you prep in a whole new way.
📫 Ask the Inbox
Q: “When I paste a full design-doc into ChatGPT and ask for a 200-word executive summary, but the reply still runs past two pages. How do I force the model to stick to the length I need without losing relevant info?”
A: Give the model hard boundaries before it starts writing and a single fallback rule if it slips:
Word-cap gate
Tell ChatGPT to stop right at the limit:
“Summarize in exactly 200 words. If the draft exceeds 200, trim the least important sentences until it fits.”Priority filter
List exactly what must survive the cut:
“Keep project goal, three success metrics, and launch date; drop implementation details.”
Put those two sentences directly under your usual Role–Task–Format–Tone prompt. The model will draft, count words, and, if it overruns, automatically prune lower-priority lines until the total is 200 words or fewer.
📌 Key takeaway: A strict word cap combined with a content-priority rule locks ChatGPT to your target length; the secret is being precise.
🤔 Got Questions?
Have a question you want answered? Email [email protected] and you just might be featured in an upcoming issue!
This newsletter is your starter kit for mastering AI with confidence. We keep things simple, show real examples, and focus on quick wins you can repeat.
Stay curious, stick with it, and watch your skills grow week by week!
Until next time :)