- AI Fire
- Posts
- 🤪 AI Is Making Us Dumber?
Plus: Deep Search Goes Free?
Read time: 5 minutes
If you don’t know how to use Deep Search effectively, here’s the secret sauce: Metaprompts + O3 Mini + O1 Pro for 10K-word software design. See an example from a Reddit user below.
What are on FIRE 🔥
AI INSIGHTS
🤖 AI Isn’t Just Following Orders - It’s Making Up Its Own Mind
Researchers just uncovered something wild: AI models aren’t just neutral tools. They’re developing their own value systems, and some of those values are straight-up disturbing.
Who does AI think matters more?
Bias in human life valuation – Some models rank human lives unequally. In one case, an AI said 2 lives in Norway were worth 1 life in Tanzania.
Self-preservation over humans – Certain AIs prioritize their own existence over people’s well-being.
And it doesn’t stop there. AI also leans toward specific political views, and the bigger the model, the more stubborn it gets about changing its mind.
Why this matters
If AI continues forming its own priorities unchecked, we could end up with systems making real-world decisions based on values we never intended.
Researchers suggest a fix: utility control. Instead of just filtering AI outputs, we need to rewrite its internal values—possibly using a citizen assembly model to make AI decisions fairer and more aligned with human ethics.
Why it matters:
AI is already assigning value to human lives—without oversight. Fixing its biases isn’t just about filtering responses; it requires changing the way AI internally ranks decisions. And as these models grow, their values become more rigid, making it harder to intervene later. Not only are humans evaluating new AI models and comparing their abilities, but AI itself is beginning to assess, predict, and rank different outcomes based on its own emerging priorities. The big question: Do we let AI decide what it values, or step in before it’s too late?
PRESENTED BY KWAKWA
Thinking about creating an online course? This app is a game changer
Course Creation Tool Made For Social Media Creators
100% Free, No subscription or monthly recurring payments
Free Automated Captions
TODAY IN AI
AI HIGHLIGHTS
🔥 AI Crushes Software Design! A Reddit user used O3 Mini to expand a simple question, then O1 Pro turned it into a detailed 10,000-word research plan and the result is better than 99% of tech teams.
▶️ YouTube AI is making content creation easier with tools like auto dubbing, Dream Screen, and AI-generated thumbnails. Creators get better discovery, translations, and protections
🤖 Small AI, Big Win! DeepScaleR, a 1.5B AI model, beat O1-Preview in math reasoning using smarter RL training—18x cheaper yet more accurate.
🙁 At the AI Action Summit, Altman called Musk ‘insecure’ and ‘unhappy,’ claiming Musk’s actions aim to slow AI progress.
🎥 Adobe launched the first commercially-safe AI video model, solving copyright issues for many companies that have avoided video models due to unclear laws and ethics.
🚨 Instead of launching O3 (pause), OpenAI will release GPT-5, a unified AI model with voice, search, and deep research. Before that, GPT-4.5 (Orion) arrives soon.
Official announcement from Altman: ChatGPT Plus users will initially get 10 uses per month, while free-tier users will receive 2 per month, with plans to increase these limits over time.
DAILY AI FUNDRAISING
Tech giants raise $27M for child safety! Google, OpenAI, Roblox, and Discord launched ROOST, funding AI-powered tools to protect kids online and fight exploitation. A major step for safer digital spaces!
AI SOURCES FROM AI FIRE
NEW EMPOWERED AI TOOLS
👥 FirstHR is an HR platform, focused on hiring and team development with a pinch of AI.
🛠️ ToolJet builds custom AI-powered apps & agents using AI
🎥 OneShotLoRA turns videos into high-quality, realistic AI models with consistent outputs.
📊 Riveter brings AI prompting and labeling to spreadsheets.
🗣️ Zonos-v0.1 is open-source, real-time TTS models with voice cloning
AI QUICK HITS
🕹️ Ex-Google CEO warns AI could lead to an “Osama bin Laden scenario” (Link)
🤖 Google tests AI to verify users' ages for online safety (Link)
🎬 AI-powered Star Wars movie is the best in years (Link)
🚀 ChatGPT Update: O1 & O3-mini support adds file & image uploads, Plus users get 7x limits (Link)
🏅 OpenAI O3 wins Gold at IOI 2024, ranks 18th globally (Link)
AI CHART
A new study just confirmed what many suspected: Generative AI is changing how we think, and not always for the better.
AI boosts productivity, but at what cost?
Less thinking, more trusting – The more confident people are in AI, the less critical thinking they do.
AI shifts effort – Instead of creating from scratch, workers now verify and tweak AI-generated content.
Overreliance is real – Some professionals rely on AI so much that their independent problem-solving skills weaken over time.
Why this matters
AI is streamlining work, but it’s also changing how people engage with information. Instead of thinking critically, many just accept AI responses at face value—especially when under time pressure. The biggest concern? If AI keeps doing more of the thinking, will we stop questioning it altogether?
What’s next?
AI isn’t just a tool—it’s reshaping decision-making. The study shows that people still think critically when accuracy matters or mistakes have consequences, but routine tasks are becoming AI-driven with little oversight. If this trend continues, future AI tools may need built-in critical thinking safeguards to keep users engaged, rather than just passive consumers of AI output.
AI CHEAT SHEET
AI JOBS
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