Yesterday Mark Zuckerberg announced a project regarding “general intelligence” for AI assistants, with Meta which is bringing together its two main research groups (FAIR e GenAI) for this purpose.
Zuckerberg commented on the project via Instagram, stating how “It has become clearer that the next generation of services requires the creation of comprehensive general intelligence“. The CEO of Meta then underlined how the opportunities are so great that this project must necessarily be open sourceso that everyone can benefit from it.
According to what was stated in The VergeZuckerberg is focusing on what is defined as AGIthat is to say General Artificial Intelligence.
This term, in reality, has a meaning that is difficult to explain today. We are talking about a hypothetical technology with a form of intelligence comparable to that of humans, useful for carrying out general tasks without the need for specific training. This form of technology, already a declared objective of OpenAIHowever, it also raises several moral doubts and more.
Zuckerberg and Sam Altman express themselves on AI and AGI, trying to throw water on the fire
Zuckberg’s announcement on Instagram makes the invention of such an artificial intelligence seem potential in the short term. In reality, at least for now, there is nothing to worry about. In fact, the time needed to develop this technology will be very long.
Earlier this week, during an interview with the World Economic Forum of Davos, the CEO of OpenAI Sam Altman stated that AI will change the world and the concept of work much less than we think today. Altman, however, underlined that the arrival of AGI is not to be placed in a distant future and will soon be a tangible reality.
Yet, contrary to the calm words of two such eminent voices, the reality seems very different. The first effects of AI on the world of work are already concrete and Google employees know something about it, with the recent layoffs. Even Elon Musk, initially at the forefront to slow down the spread of this new technology, then joined his “colleagues” by proposing Grok.
Zuckerberg himself recently stated that Llama 3 is in the training phase and that Meta is being structured with adequate GPUs to train AI models at a massive level.