▾OpenAI unveiled the revolutionary model o3. But we'll have to wait for the launch
OpenAI concluded its Christmas event "12 Days of OpenAI" by announcing the revolutionary model o3 and its smaller version o3-mini. The new model promises significant improvements in reasoning and solving complex tasks. For now, it will only be available to safety researchers.
The company OpenAI surprised the world by announcing the new model o3 and its smaller version o3-mini, successors to the successful o1 series. However, it remains under wraps and will only be available to AI safety researchers. This announcement marks the culmination of the "12 Days of OpenAI" Christmas event, which brought a host of groundbreaking news.
Why is o3 so significant?
The new model represents a significant leap in the capabilities of artificial intelligence. Unlike its predecessors, it offers more advanced reasoning and complex task processing capabilities. OpenAI is keeping specific specifications under wraps for now, but it's expected that o3 will surpass current models in logical thinking and problem-solving.
Safety first
OpenAI is taking a cautious approach by first providing access only to safety researchers. This step reflects the growing emphasis on responsible AI development and the effort to prevent potential risks. Researchers will test the model before it is widely deployed and check for any potential security gaps.
Although the exact release date is unknown, it's expected that o3 could be available to a broader public during 2025. OpenAI will likely opt for a phased release, similar to its previous models.
In addition to o3, the company introduced several interesting updates. Among the most significant are real-time image processing, an improved ChatGPT with search functionality, and a Christmas surprise in the form of Santa Claus's voice for ChatGPT.
What consumes the most data on mobile? YouTube, Spotify, Netflix and other apps under scrutiny
Mobile data usage can sometimes be unpleasantly surprising. The most are consumed by video streaming and music services – how much data does YouTube, Netflix or Spotify use? And what about other apps like social networks, maps or video calls? Check out the overview of the biggest data guzzlers and find out how to keep mobile data under control.
How to find out who is connected to your Wi-Fi – and can you easily disconnect them?
Is your internet slowing down or do you suspect that someone unauthorized is connected to your Wi-Fi? We advise you on how to find out who is connected to your Wi-Fi, and what to do if you find an intruder.
Artificial intelligence is not omnipotent. Where does the human still have the advantage?
Artificial intelligence can write articles, analyze data, and generate videos. However, there are skills and professions that this technology cannot (yet) replace. In the article, we will look at where humans still have the upper hand, why AI needs human supervision, and which abilities will become increasingly valuable to develop in the digital world.
What internet speed is sufficient? Comparison for streaming, gaming, work, and TV
Does your video keep buffering or game lag, even though you pay for 'fast internet'? Perhaps you simply don't have the right one for what you do at home. What internet speed do you need for watching TV, streaming movies, working from home, or online gaming? Find a clear comparison in the article.
Phishing in the Age of AI: How to Recognize a Fraudulent Email When It No Longer Looks Suspicious
Phishing attacks often appear credible and convincing today. Thanks to artificial intelligence, they sound personal, mimic well-known companies, and can be hard to detect. In this article, you'll learn how modern phishing works, why you should watch out for it, and how you can effectively protect yourself from it.
Passkeys: what it is and why you'll soon stop logging in with a password?
Forgotten passwords, recycled combinations, and fear of data leaks. Traditional login methods have their weak points. A new way to verify identity, known as passkeys, aims to change that. Without the need to remember anything and with a focus on security. How do they work, who is already using them today, and why are they becoming the new standard?