- AI Trailblazers - We moved to MailChimp
- Posts
- OpenAI releases o1, its first model with ‘reasoning’ abilities
OpenAI releases o1, its first model with ‘reasoning’ abilities
+ OpenAI’s ChatGPT surpasses 11 million paying subscribers, generating billions in revenue
Your team's new best friend: Atlassian
Imagine if your tools could think a little more like you. Atlassian Intelligence brings AI smarts to your favorite Atlassian apps, taking care of the busywork and leaving you with more time to focus on the good stuff. From instant insights to seamless automation, it’s like having a supercharged assistant who’s always one step ahead. Let your team do what they do best—minus the hassle.
Latest headlines 👇
Think it’s impossible to change someone’s mind about conspiracy theories? New research says otherwise. A study found that chatting with an AI tool, "DebunkBot," made conspiracy believers rethink their stance, reducing belief by an average of 20%. By using personalized, fact-based responses, the AI helped chip away at entrenched ideas like moon landing hoaxes and microchip-laden vaccines. The experiment challenges the long-held belief that once people fall down the conspiracy rabbit hole, they’re stuck there. Read more
OpenAI just launched its latest "Strawberry" series of AI models, designed to handle complex problems with advanced reasoning skills. Dubbed o1 and o1-mini, these models are capable of breaking down tough tasks in science, coding, and math with impressive accuracy. The o1 model even scored 83% on the International Mathematics Olympiad qualifying exam—a huge leap from previous models. By automating "chain-of-thought" reasoning, OpenAI's latest models think through problems like a human would, refining their approach along the way. Read more
Top executives from Nvidia, OpenAI, Anthropic, and Google gathered at the White House to discuss AI’s growing energy and data infrastructure needs. The conversation, which included leaders from Microsoft, Amazon, and U.S. power companies, focused on AI’s energy usage, semiconductor manufacturing, and data center capacity. With AI's rapid expansion, Nvidia CEO Jensen Huang emphasized the importance of public-private collaboration to meet energy demands efficiently. The White House has launched a new task force to ensure the U.S. remains at the forefront of AI innovation while addressing these challenges. Read more
OpenAI's ChatGPT has reached over 11 million paying subscribers, including 1 million on higher-priced business plans, according to COO Brad Lightcap. This milestone is driving an estimated $225 million in monthly revenue, putting the company on pace to generate $4 billion annually. While ChatGPT’s growth is unprecedented, it comes at a steep cost, with the company facing significant financial losses. OpenAI is exploring new revenue streams, including a potential $7 billion investment and evolving its business model to sustain its rapid expansion. Read more
Meta has revealed that every public Facebook and Instagram post, dating back to 2007, has been used to train its AI models. During a government inquiry, Meta’s privacy director confirmed that unless posts were specifically set to private, they were included in the data collection. While European users can opt out due to privacy laws, the rest of the world, including Australia, has no such option. This raises significant concerns about data use, especially for content posted by minors or without user awareness at the time. Read more