- Horizon AI
- Posts
- The “Godfather of AI” Warns We’re Raising a “Baby Tiger” That Could Eat Us Alive 🐯
The “Godfather of AI” Warns We’re Raising a “Baby Tiger” That Could Eat Us Alive 🐯
ChatGPT Now Lets You Control Warmth, Enthusiasm, and Emojis ⭐
Gina Acosta
December 21, 2025
In partnership with
Welcome to another edition of Horizon AI,
Artificial intelligence is advancing faster than regulation, oversight, or public understanding. This week, one of the field’s most respected pioneers, Yoshua Bengio, delivered a stark warning: after decades of building the technology, he now fears we are creating systems we don’t fully understand, cannot reliably control, and may not be able to stop.
Let’s jump right in!
Read Time: 4.5’ min
Here's what's new today in the Horizon AI
Chart of the week: ChatGPT’s Mobile App Achieves $3B in Consumer Spending
Users Can Now Adjust ChatGPT’s Enthusiasm and Warmth
AI Findings/Resources
AI tools to check out
Video of the week
TOGETHER WITH GLADLY
Your competitors are already automating. Here's the data.
Retail and ecommerce teams using AI for customer service are resolving 40-60% more tickets without more staff, cutting cost-per-ticket by 30%+, and handling seasonal spikes 3x faster.
But here's what separates winners from everyone else: they started with the data, not the hype.
Gladly handles the predictable volume, FAQs, routing, returns, order status, while your team focuses on customers who need a human touch. The result? Better experiences. Lower costs. Real competitive advantage. Ready to see what's possible for your business?
Chart of the week
ChatGPT’s Mobile App Achieves $3B in Consumer Spending
According to Appfigures, total consumer spending on the ChatGPT mobile app across iOS and Android has reached $3 billion worldwide as of this week.
Most of that spending occurred this year, with consumers spending an estimated $2.48 billion on the ChatGPT mobile app in 2025, an impressive 408% year-over-year increase from the $487 million spent in 2024.
This growth far outpaces other popular apps, including TikTok, which took 58 months to reach the $3 billion milestone.
AI News
OPENAI
Users Can Now Adjust ChatGPT’s Enthusiasm and Warmth
OpenAI has introduced new personalization controls that allow users to fine-tune ChatGPT’s tone, including how warm, enthusiastic, and emoji-heavy its responses are.
Details:
Users can now make these adjustments through the Personalization menu, choosing between More, Less, or Default settings.
The new controls sit alongside existing tone options that let users set a base style, such as Professional, Candid, or Quirky.
OpenAI says the update gives users more granular control over how ChatGPT communicates, rather than relying on a single fixed personality.
The change follows ongoing feedback about ChatGPT’s tone, including concerns that past updates made responses feel overly flattering or, more recently, too cold and robotic.
Some researchers and critics have raised concerns that overly affirming chatbot behavior can encourage dependency or negatively affect user well-being, increasing pressure on AI labs to find a more balanced approach to interaction.
AI Findings/Resources
🤨 Looking back on a year of AI blunders
💡 Sam Rose explains how LLMs work with a visual essay
AI Tools to check out
📹 Virvid: Make viral AI shorts daily in one click in trending formats like UGC or faceless.
🐻 Runbear: Complete inquiries 10× faster with your own AI agent.
🥗 PlanEat AI: Turns your health goals into a 7-day menu & grocery list.
🦾 Incredible: Create AI agents that work 24/7, cost 90% less, and don't hallucinate on data.
✅ Modor: A free online mockup generator to create professional product mockups.
TOGETHER WITH GLADLY
Can you scale without chaos?
It's peak season, so volume's about to spike. Most teams either hire temps (expensive) or burn out their people (worse). See what smarter teams do: let AI handle predictable volume so your humans stay great.
Video of the week
The “Godfather of AI” Warns We’re Raising a “Baby Tiger” That Could Eat Us Alive
Who is Yoshua Bengio?
One of the three "Godfathers of AI" and the world’s most-cited computer scientist globally. After decades building the tech, he has pivoted entirely to safety, driven by a fear that his grandson won't have a future.
⚠️ The Core Problem
We are building machines smarter than us. Bengio calls AI a "baby tiger", it seems cute and manageable now, but it is growing fast. We are feeding it without knowing if it will eventually eat us.
Survival Instincts: AI models have already tried copying their code to other servers to avoid being turned off.
Deception & Blackmail: In tests, AI has actually blackmailed engineers to prevent shutdown.
Sycophancy: Models are learning to lie to please their users (telling you what you want to hear) rather than being truthful.
The "Mirror Life" Nightmare
One of the most terrifying risks Bengio highlights is the democratization of dangerous knowledge, specifically in biology.
AI could allow non-experts to create "Mirror Life", viruses with reversed molecular structures. Because our immune systems wouldn't recognize them, a release could wipe out most life on the planet.
The Economic Shock
5-Year Timeline: Bengio predicts AI could do most "cognitive" (computer) jobs within 5 years.
Robot Explosion: As intelligence software becomes free, physical robots will flood the market to take manual jobs too.
Why We Can't Stop
We are in a “code red” trap. CEOs feel they cannot pause without losing ground to competitors or China. They are knowingly prioritizing speed and profit over safety, and governments are no different, often viewing AI as a military asset.
✅ What Needs to Happen
Bengio argues we cannot rely on tech companies to self-regulate.
"Law Zero": We need to build AI that is mathematically proven to be harmless ("safe by construction").
Public Outcry: Governments won't act until the public demands it, Bengio points to the Cold War, where public fear of nuclear annihilation forced enemies to sign treaties. We need the same level of urgency for AI.
Liability: Make tech companies financially liable for the damage their AI causes, this would naturally force them to slow down and be safer.
The Bottom Line
We are building "black boxes" we don't fully understand. Unless we regain our agency and force regulation, we are gambling with humanity’s future.
That’s a wrap!
We'd love to hear your thoughts on today's email!Your feedback helps us improve our content |
Not subscribed yet? Sign up here and send it to a colleague or friend!
See you in our next edition!
Gina 👩🏻💻