- FryAI
- Posts
- DoomsDay clock ticks down...
DoomsDay clock ticks down...

Good morning! We’ve cooked up some juicy news, tools, and insights for you to feast on today. 😋
🤯 MYSTERY AI LINK 🤯
(The mystery link can lead to ANYTHING AI-related: tools, memes, articles, videos, and more…)
Today’s Menu
Appetizer: Rapid AI development is “a very risky gamble” 😳
Entrée: Did DeepSeek steal data from OpenAI? 👀
Dessert: Alibaba releases new model to rival DeepSeek 🦾
🔨 AI TOOLS OF THE DAY
🤵♂️ HeyMilo: Conduct interviews with AI voice. → Check it out
🧍♀️ PostureWatch: Fix your posture. → Check it out
🤨 Truth or Dare AI: Use AI to enhance your game of Truth or Dare. → Check it out
RAPID AI DEVELOPMENT IS “A VERY RISKY GAMBLE” 😳
Some personal news: After four years working on safety across @OpenAI, I left in mid-November. It was a wild ride with lots of chapters - dangerous capability evals, agent safety/control, AGI and online identity, etc. - and I'll miss many parts of it.
— Steven Adler (@sjgadler)
5:20 PM • Jan 27, 2025
Will AI destroy humanity? 🤖
What’s up? A former safety researcher at OpenAI, Steven Adler, has voiced concerns over the rapid pace of AI development, calling it a “very risky gamble.”
What did he say? In a series of posts on X (above), Adler said he was “pretty terrified” by how quickly AI is advancing, questioning whether humanity is prepared for the consequences. Adler, who left OpenAI in November, worked on safety research for both short-term AI products and long-term, speculative AI systems. He warned that no company has solved the problem of AI alignment—ensuring that AI follows human values—and that the race to develop artificial general intelligence (AGI) is making it harder to do so. He stated, “Even if a lab truly wants to develop AGI responsibly, others can still cut corners to catch up, maybe disastrously.” Adler insists that without strong safety regulations, the industry may be moving too fast to prevent potential disaster. These fears were echoed by the famous DoomsDay Clock, which has been set at 89 seconds to midnight, the closest it has ever been.
“I’m pretty terrified by the pace of AI development these days. When I think about where I’ll raise a future family, or how much to save for retirement, I can’t help but wonder: will humanity even make it to that point?”
DID DEEPSEEK STEAL DATA FROM OPENAI? 👀
Two guys were arrested for stealing a calendar. They each got six months. 📆
What’s going on? OpenAI is working with the U.S. government to investigate whether DeepSeek illegally used ChatGPT outputs to train its new R1 model.
Want the background? This week, a Chinese startup called DeepSeek, founded just 20 months ago, released a model called R1 that purportedly outperforms GPT-4o on many tasks. The catch? DeepSeek claims to have only spent $6 million to build the model and used much less powerful GPUs than those available to OpenAI. To put this in perspective, GPT-4o is estimated to have cost at least $100 million for OpenAI to develop, over 16x that of DeepSeek R1.
What is being investigated? While some are praising DeepSeek for creating such a great model with little resources, OpenAI remains skeptical. Many experts are questioning whether DeepSeek truly built its model from scratch, with some suggesting it may have relied on OpenAI’s work. White House AI advisor David Sacks claims DeepSeek may have extracted knowledge from OpenAI’s models through a process called distillation—a machine learning technique that transfers knowledge from a large model to a smaller model.
“It is unclear whether DeepSeek really trained its models from scratch. OpenAI has stated that they believe DeepSeek may have misappropriated large amounts of data from them. If this is the case, then the claims about training the model very cheaply are deceptive. Until someone replicates the training approach, we won’t know for sure whether such cost-efficient training is really possible.”
ALIBABA RELEASES NEW MODEL TO RIVAL DEEPSEEK 🦾
I said, “Alibaba!” … and a new AI model appeared. 🪄
What’s up? Alibaba has unveiled Qwen 2.5 Max AI model, claiming it outperforms OpenAI’s GPT-4o, Meta’s Llama-3.1, and DeepSeek-V3.
Want the details? DeepSeek has disrupted the market with its low-cost, high-performance AI models. Its DeepSeek-V3 and R1 models are said to rival GPT-4o but at a fraction of the cost. This shook Silicon Valley, leading to stock drops and prompting rivals like Alibaba, Tencent, and Baidu to accelerate AI development. The AI race has also triggered a price war. DeepSeek’s aggressive pricing has forced Alibaba to slash costs by up to 97% on some models.
TASTE-TEST THURSDAY 🍽️
Do you think other countries will catch the U.S. in terms of AI innovation?(Leave a comment explaining your answer, and we might feature it tomorrow with the results.) |
HAS AI REACHED SINGULARITY? CHECK OUT THE FRY METER BELOW:
What do ya think of this latest newsletter? |