- AI Fire
- Posts
- 🚫 DO NOT Vent to ChatGPT!
🚫 DO NOT Vent to ChatGPT!
3 New AI Agents You Haven't Heard Of

Read time: 5 minutes
AI gadget turns your dreams into trippy movies. Microsoft’s new AI beats real humans at diagnosis. But even when AI says something nice, we don’t really feel it. Why?
What are on FIRE 🔥
LEARNING PARTNER AIRCAMPUS
No Code. No Overwhelm. Just Drag, Drop, Deploy
Imagine building a dream team -
Not with resumes or interviews. But with building blocks.
📦 Drop in an Inbox Agent to handle messages.
🧠 Add a Content Bot that writes while you sleep.
🤝 Connect a Follow-Up Pro to close every lead.
All flowing together - automated, effortless, alive.
No tech background. No friction. Just creation.
⚙️ This isn’t coding. This is designing your operations.
Like a canvas - except every box gets things done.
🔘 Reserve Your Spot — Build in 3 Hours
AI INSIGHTS
What if you could watch your dream like a short film? A Dutch design studio just made that possible sort of. Meet Dream Recorder, a tiny AI device that turns your sleepy, half-remembered dreams into grainy video replays.
Here’s how it works: You wake up → You push a button → You speak your dream out loud → You get a surreal, low-res video in your style, your language, your imagination.
Blurry castles, floating rooftops, flowers made of light, whatever your brain cooked up during REM.
The code, the hardware design, the instructions, all on GitHub. You build it yourself.
→ Total build cost is around €285, plus a few cents in AI fees per dream (~$0.01–$0.14 depending on quality).
There’s space for 7 dreams, one for each night of the week. And you can even add style prompts to shape the visuals.
But here’s what really makes this special: The intent behind it. Modem didn’t build this to boost screen time or collect user data. In fact, it’s designed to do the opposite.
→ Modem calls this "ambient computing" - a nod to early ideas from Xerox PARC's Mark Weiser: calm technology that fades into the background instead of fighting for attention.
Why It Matters: While most new devices want to automate, monetize, or optimize, Dream Recorder just… listens. Quietly. Patiently. It asks, what if AI could help you understand yourself a little better? They’ve even set a sunset date for their studio: 2030. No IPO. No infinite roadmap. Just thoughtful, time-boxed invention.
PRESENTED BY CODECADEMY
Looking to stand out in a competitive job market? Codecademy can help you build in-demand technical skills and prepare for top industry certification tests from AWS, Microsoft, CompTIA, and more.
AI Fire readers can save 15% on a year of Codecademy Pro and get access to:
600+ interactive courses
Step-by-step career guidance
Industry certification prep
Try hands-on learning and get ready to unlock new career opportunities — just use code SKILLUP15 at checkout for 15% off
TODAY IN AI
AI HIGHLIGHTS
🎓 When ChatGPT = CheatGPT, 90% of college students used it on assignments. We’re sure you saw the video of the UCLA grad student flexing his use of ChatGPT. If not, here it is.
💼 Meta hired 4 more OpenAI researchers (now eight in all), shedding new light on the CEO’s “secret list” of AI experts. P.S: at least one said they did not receive $100M.
💀 Friendly reminder: do NOT trust ChatGPT with your mental health; multiple people have been involuntarily committed due to psychosis triggered by chatting with GPT.
📱 Did these viral, AI-generated cell phone videos fool you? These hyper-realistic clips look like they were shot with phones or cameras. “We are all so getting scammed when we’re old”
🩺 Microsoft's new AI diagnoses 4 times more accurately than human doctors while cutting testing costs by about 20%. It uses a chain-of-debate model with multiple top AI models.
🖼 Canva now works directly inside ChatGPT, you can edit designs without leaving your chat. It’s the first design connector for ChatGPT & supports one-click integration.
💰 AI Daily Fundraising: Tandem Health secured $50 million to enhance its AI tool that reduces clinician burnout, marking a significant step in transforming healthcare operations across Europe.
AI SOURCES FROM AI FIRE
NEW EMPOWERED AI TOOLS
🖼️ Pokecut uses AI to create photos with just a few click or a prompt for free.
🗣️ Jotform creates AI presentations that talk, listen and answers from files.
🎨 Picsart Ignite 2.0 instantly generates branded assets, ads, videos, fonts,…
💻 DemoDazzle gets an interactive assistant that looks & sounds like you.
🧩 Tabl 1.0 brings Figma-like collaborative experience to any app, on any tab.
AI QUICK HITS
🔥 Meta launched Superintelligence Labs with 11 top researchers poached.
📚 Google expanded access to Gemini for Education with 30+ AI features.
🔮 A scientist said we will reach the AI singularity within 20 years, by 2045.
📅 Elon Musk said that Grok 4 is planned for a release ‘'just after July 4”.
👨💻 Cursor launched a web app to manage a network of AI coding agents.
AI CHART
Turns out, even when AI says the right thing, people still want to hear it from an actual person. A new study dropped this week, 9 experiments, 6,000+ people, one surprising truth: we still prefer empathy from humans.
Researchers gave participants empathic replies. Sometimes they said it came from a person. Other times, from a chatbot. But all generated by AI.
What happened → Participants rated the “human” responses as warmer, more supportive, and way more emotionally satisfying.
The kicker: People were willing to wait days or even weeks just to get that same response but from a human!?
But wait, it gets weirder. When people thought a human might have used AI to “help” write their message? Boom trust dropped.
Even the idea that a message was “AI-touched” made it feel less sincere. So what’s going on here? Here’s what this tells us:
Same words, different impact: It's not just what you say, it's who people think said it.
Perceived authenticity is everything: A message feels more real when we think it came from someone who took time and emotional effort.
AI’s emotional ceiling is real: It can scale support, but not deep connection.
As LLMs write more of our emails, DMs, and apologies, the risk is clear: we may stop believing anyone means what they say.
AI can simulate empathy. But it can’t feel it. That gap might be the most important one we don’t want to close. So yeah, next time you're tempted to “run it through ChatGPT”… maybe just say it like you mean it. People can tell.
We read your emails, comments, and poll replies daily
How would you rate today’s newsletter?Your feedback helps us create the best newsletter possible |
Hit reply and say Hello – we'd love to hear from you!
Like what you're reading? Forward it to friends, and they can sign up here.
Cheers,
The AI Fire Team
Reply