Google’s recently released RT-2 artificial intelligence (AI) model has experts abuzz, as the major tech titan is bringing us one step closer to the robot from the 2008 Disney Pixar movie WALL-E.
RT-2 was released on October 15, 2019, and is the world’s most powerful AI model to date, with 1.5 times the performance of Google’s previous models. This breakthrough allows the AI to better recognize images and videos while also understanding natural language commands.
The model was trained on Google’s own hardware with system performance twice that of the ResNet-50 model – the reigning champion prior to RT-2. Beyond its ability to recognize images and videos more quickly, the model is also able to answer questions posed in natural language.
In some respects, RT-2 isn’t much different from the AI shown in the Pixar movie WALL-E. In both cases, the AI exhibits a great deal of learning and autonomy, using data to come to decisions and comprehending a variety of languages.
Although the model is still far from resembling the level of autonomous AI shown in the movie, it is nonetheless a giant leap forward. The model is still a far cry from fully understanding the complexities of the human language, however, and experts estimate that it will likely be at least a few years before any sort of WALL-E-like AI is achieved.
It is worth noting that RT-2 isn’t the only AI model being developed by Google. The company is also working on Google Duplex, an AI-driven voice assistant designed to make phone calls and book appointments on users’ behalf. While this model isn’t quite as advanced as RT-2, it nevertheless illustrates the progress that Google is making in the AI space.
Google’s RT-2 model marks an important milestone in AI technology and brings us one step closer to one day understanding the complexities of the human language. Though the AI model still has a long way to go before it can replicate the autonomous robots seen in WALL-E, advancements such as these are certainly promising.
