- In The Loop AI
- Posts
- o3 OpenAI model hallucinates more than previous models
o3 OpenAI model hallucinates more than previous models
ALSO: New AI startup to replace human workers
Algot Persson
April 19, 2025
In partnership with
In Today’s Issue:
o3 OpenAI model hallucinates more than previous models
New AI startup to replace human workers
Read time: 3 minutes
o3 OpenAI model hallucinates more than previous models
OpenAI’s latest language model, o3, is reportedly producing more hallucinations than earlier versions, according to developers testing the model through OpenAI’s API and various leaked reports. The model, believed to be a next-generation successor to GPT-4, has been quietly rolled out to some users under the name “gpt-4o.”
o3 has been praised for its speed and responsiveness, showing clear improvements in conversation flow and performance. However, developers say it tends to generate more inaccurate or misleading information, especially in areas that require precision, such as law, science, or programming. Some testers have noted that the model sounds more confident when delivering these incorrect answers, which makes the hallucinations harder to detect. OpenAI has not publicly confirmed or commented on the issue.
Sam Altman
Run Your Sales on Autopilot
Increase the output of your sales team without buying more tools or hiring new SDRs. Onboard Agent Frank, Salesforge’s AI SDR, to fully automate prospecting, personalized outreach and booking meetings while your team focuses on closing deals. Get a personalised email from Agent Frank to see his work in action:
New AI startup to replace human workers
A new startup by a leading AI researcher announced itself with a brash goal in mind: to replace all human workers. The startup, Mechanize Inc. was announced on X:
I'm starting a new company: Mechanize.
Mechanize will build virtual work environments, benchmarks, and training data to enable the full automation of all work.
We're hiring: [email protected].
— Tamay Besiroglu (@tamaybes)
4:26 PM • Apr 17, 2025
The full announcement was met with criticism from tech professionals. Replies from the reply section called the goal “reckless” and warned of the social risks of aggressively automating human labor. Others questioned whether the technology is even close to being capable of such a transformation. Despite the backlash, Mechanize has already sparked strong interest from investors drawn to its ambitious pitch and early momentum. AI maximalism in 2025 is synonymous with replacing all human labor. "AI safety" being the #1 priority has taken the backseat for now.
Which company has the best AI by the end of April?:
How was today's newsletter? |
Reply