Next Article
Google DeepMind wants to build human-like AI by making models bigger
Technology
Google DeepMind is going all-in on creating artificial general intelligence (AGI)—basically, an AI that can think and learn like people do.
At the recent AI+ Summit, CEO Demis Hassabis said their main strategy is to scale up current AI models, hoping this will lay the groundwork for true AGI.
The big picture (and some debate)
AGI is still just an idea for now, but it's driving huge investments and a race to build smarter machines.
While DeepMind believes "bigger is better," not everyone's convinced.
Experts like Yann LeCun warn that simply adding more data and computing power might not solve everything—he's betting on new ways for AI to learn from real-world experiences instead.