Earlier today, Mark Zuckerberg participated His vision is determined to build “Superinteigence”. In the note, Zuckerberg hinted that the stronger pursuit of artificial intelligence may require the company to be more selective in what opens the sources.
Quoting “safety concerns”, he wrote that dead will need to be “strict” about such decisions. The line has emerged to many like Zuckerberg – who once said.“In reference to closed platforms – it made the Meta strategy of artificial intelligence.
During the profit call in the second quarter of Mita, Zuckerberg also acknowledged that there may be a shift, although he reduced the importance of this. Here is what he said when asked if his thinking had changed.
I don’t think our thinking has changed in particular in this. We have always opened some of our models and do not open everything we have done. Therefore, I expect to continue to produce and share open source models. I also think there are some directions that play. One of them is that we get so large models that they are not a practical use of many other people, so we are dealing with whether it is regular or it is useful to share this, or if this is primarily helps competitors or something like that. So I think there is, there is this anxiety.
Obviously, as you approach the real inspiration, I think there is a completely different group of safety concerns that I think we need to take it seriously, I wrote about it in my note this morning. But I think the bottom line is that I expect to continue in open work. I expect us to continue to be two leaders there, and I also expect to continue not to open everything we do, which is a continuation of the kind of what we were, what we were working on.
This is significantly different from what he wrote almost a year ago in A. Entitled “Open Source Ai is the path forward.” In it, even a longer note, he said that the open source is very important for both Meta and developers.
“People often ask if you are concerned about giving up a technical feature through open sources Lama, but I think this misses the big picture,” he wrote. “I expect the development of artificial intelligence to continue to compete very much, which means that open sources do not give any specific model an enormous advantage over the best models in that time stage.”
He also argued that the open source is Safety. “There is a continuous discussion about the safety of the open source AI models, and my opinion is that an open source AI will be safer than alternatives. As long as everyone has access to similar generations of models – which are enhanced by the open source – then governments and institutions that have more reckoning resources, you will be able to verify bad actors with the least pressure.”
To be clear, Zuckerberg said that the company will continue to open some of its business. But it seems that it puts the foundation for a future in which Meta is less open.
https://s.yimg.com/ny/api/res/1.2/9ZtPWE4HSdGNRQEXo8bmsA–/YXBwaWQ9aGlnaGxhbmRlcjt3PTEyMDA7aD04MDA-/https://s.yimg.com/os/creatr-uploaded-images/2023-12/5f1d5d80-a46b-11ee-bbbf-75882e35cf7c
Source link