- AI Fire
- Posts
- 👥 Is Every AI Playing Copycat?
Read time: 5 minutes
Are AIs Just Trend-Chasing Now? One AI drops a cool feature, and suddenly the whole class has it. Memory features, workspaces, deep research, reasoning, “Agent-style”,… Coincidence? Strategy? It’s the iPhone vs. Android vs. Pixel war…
What are on FIRE 🔥
LEARNING PARTNER: AIRCAMPUS
Learn AI agents & automation to become AI-proof | Unlock New Revenue | No Coding!
Join this power-packed 3-hour AI Masterclass and learn how to Automate Workflows, Supercharge Productivity, and Unlock New Revenue Streams with cutting-edge AI tools used by top professionals and Master AI Agents.
Zero Coding Required!
If you’re still figuring out how to use AI in your career, you’re already behind.
What You’ll Gain in Just 3 Hours:
✅ Automate repeated and boring tasks
✅ Master 5+ AI tools & hacks to leverage AI
✅ Learn Agentic AI, Workflow and Implementation
✅ AI Agents, sub-agents, and tools design & architecture
🔥 BONUS: Get exclusive Gen AI templates & workflows
📆 Date: 19th April, Friday | 🕒 Time: 10 AM EST
🎟️ FREE for the first 50 professionals (Worth $299)
AI INSIGHTS
OpenAI’s newest model, o3, has officially crossed a threshold: it scored 136 on the Mensa Norway IQ test, placing it higher than 98% of the human population. It’s a 30-point leap from earlier models.
Some experts are calling it the end of the 300,000-year human IQ monopoly. And they might not be joking.
Key Highlights:
Scored 136, which is above the Mensa entry threshold, and puts it in the top 2% of human intelligence.
Achieved using a 7-run average, not a one-off lucky result. o3 outscored models like GPT-4o, Claude Opus, and Gemini Advanced.
This performance gap underscores how resource-intensive top-tier reasoning models are.
With companies like OpenAI not disclosing fine-tuning data or internal benchmarks, third-party testers are becoming the de facto evaluators for the public and academic community.
Signals for AGI and Agent Development → While IQ tests aren’t the full picture, models like o3 scoring at this level show clear advancement toward human-level reasoning in short contexts.
Why It Matters: Pour one out for our 300,000-year reign as the smartest species on the planet was a great run. And raw intelligence is going up, fast. It’s now a race between intelligence and control and humans are no longer guaranteed to be ahead.
🎁 Today's Trivia - Vote, Learn & Win!
Get a 3-month membership at AI Fire Academy (500+ AI Workflows, AI Tutorials, AI Case Studies) just by answering the poll.
Which AI model has the SECOND highest IQ score right now? |
IN PARTNERSHIP WITH GENEVA TOURISM
Geneva, Switzerland is famous for luxury watches and international summits, but it’s also a hub of scientific innovation. At CERN, you’ll find the world’s largest particle accelerator and the cutting-edge CERN Science Gateway. Designed by renowned architect Renzo Piano, the Science Gateway offers hands-on exhibits, immersive VR, and live demos that bring the mysteries of the universe to life. Whether exploring particle physics or experiencing the Big Bang, it’s an unforgettable journey for visitors of all ages. Next time you're in Geneva, go beyond the scenic lake and charming Old Town—dive into the future at CERN’s Science Gateway.
TODAY IN AI
AI HIGHLIGHTS
🤩 OpenAI launched a new pricing option called Flex processing for its o3 and o4-mini AI models with new ID verification rules, and it’s 50% cheaper than normal pricing.
🎓 Google One AI Premium - including 2TB of storage, Veo 2, Gemini 2.5 Pro & other tools - is Free for college students until Spring 2026. Sign up before the deal expires.
💸 Every “thank you” or “please” costs OpenAI GPT Tens of millions of Dollars annually. Sam Altman quips that it's money 'well spent' (adding 'you never know').
⚙️ Gemini 2.5 Flash with “thinking budget” rolling out to devs, Gemini app. It can basically "morph" itself to flexible modes using “thinking sliders”.
🛑 Google says LLMs.txt - a proposed standard for showing website content to AI agents and crawlers - is comparable to keywords Meta tag. But why don’t OpenAI, Google, or Anthropic use it?
🤖 xAI's new memory feature in Grok remembers past chats for more personalized interactions. But it will come with “Incognito Mode” or “Blank Slate”?
💰 AI Daily Fundraising: Cybersecurity startup Exaforce raised $75M to fix false alerts using AI. Backed by Khosla Ventures, Mayfield, and Thomvest, the 2-year-old firm didn’t reveal its valuation.
AI SOURCES FROM AI FIRE
🎉 Finally, we’re back with some love for our smartest readers!
You’ve been voting, guessing, and proving once again that this community is full of brainpower. Now it’s time to celebrate the trivia legends of February & March 2025💥💥
fr*****[email protected]
ja*****[email protected]
xe*****[email protected]
el*****[email protected]
re*****[email protected]
og*****[email protected]
ry*****[email protected]
to*****@csun.edu
(And yes, April winners coming soon... stay sharp 👀)
NEW EMPOWERED AI TOOLS
👩💻 OpenAI Codex CLI brings GPT‑4.1 power right into your terminal.
🖱️ CursorFlow brings live trails, click pops, and keystroke overlays to your laptop.
🎬 Kling AI 2.0 adds, swaps, or delete any elements in a video.
🔍 Claude Research operates agentically, conducting multiple searches.
🔮 ChatGPT just got mind-blowing computer vision powers like in the movies.
AI QUICK HITS
🤖 Wikipedia shows its AI dataset due to being sick of scraper bots.
🚫 EU has officially banned AI agents from joining its meetings.
🔥 ChatGPT's pet-to-human trick takes TikTok by storm with millions views.
🛡️ OpenAI will phase out GPT-4.5 yet just releases in late February.
🐶 AI robot dog is learning to 'live' like a human using physical agentic AI.
AI CHART
The report came three weeks after the public release of Gemini 2.5 Pro (in “preview” mode). It provides useful - and unflattering, at times - info that companies don’t always widely advertise about their AI. It’s too sparse to assess real-world risk or safety readiness.
This report doesn’t mention Google’s own FSF designed to catch future “severe harm” risks.
No mention of tests for dangerous capabilities.
Google says those are handled in a separate audit, but that’s not public?
We see an interesting thing here. Google only publishes safety reports after models leave the "experimental" stage while Anthropic & OpenAI often share safety insights pre-release.
So, we wonder if Google, OpenAI, and others are so under immense pressure to release models fast that they starts “race to the bottom” trend. And because of this delay, OpenAI and Meta are also backsliding on safety transparency.
AI JOBS
Tiposi: AI/ML Engineer (Image Generation) ($100,000 - $120,000 Annual)
Meta: Research Scientist Intern, Monetization Generative AI ($7,800/month to $11,293/month + Benefits)
Next-U Education: AI Product Design Part-time ($400 – $500 a day)
Ztek Consulting: Salesforce Gen AI Architect Onsite or Remote ($80 – $90/h)
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