From a general perspective, have we already reached a critical point regarding the development of Artificial Intelligence? In the case of environmental preservation, we know that for now, if humanity becomes aware and decides to implement sustainability actions, we can still have this planet called Earth functioning and providing for all humans on it for a long time. Now, making an analogy with AI technology, will we have a similar critical point?
The development of this technology has advanced every day, causing this question to be constantly asked by people. Based on the Turing Test, we will see if these AIs are already reaching a level close to SkyNet, or if we still have time, as in the matter of environmental preservation, to take the correct actions to prevent the emergence of some type of AI with consciousness that could harm the human race.
I'll be much appreaciated if you Votes, Reblogs or give any Tips. Thanks and have good reading! 🍀
Have We Reached the Critical Point of AI Yet?
To contextualize the analysis and reflection, it's worth understanding what the Turing test is. The Turing Test was created by Alan Turing in 1949/1950. During this wartime period, the theoretical advancement of computing was very significant because there was a war in the area of technology where both sides sought to decode each other's messages. The Turing Test begins with the question: "Can machines think?" At the time, computers and large machines were new technologies; our current form of Artificial Intelligence did not yet exist, so the most commonly used term was "machines," but we can consider them as AIs or robots. The Turing Test was a way to determine whether a machine would be intelligent or not. In this test, there would be three participants, as shown in the figure above. The first player, A, would be the machine attempting to impersonate a human; participant B, a normal human acting like a normal human, answering questions normally; and finally, player C, a tester who would ask the questions and differentiate a machine from a human.
At that time, computers were not capable of accurately simulating human behavior, and several predecessors of current AI models failed the Turing Test. Turing believed that if a machine, AI, or robot could impersonate a human, making the person asking the questions believe it was a human, then that machine could be considered "Intelligent." It's not hard to imagine that today AIs are capable not only of solving the Turing test, which would be done through text, but are reaching the point of being able to imitate human voices and also generate realistic videos. For those watching content without paying close attention, depending on the video and the adjustments made, finding the difference between a video made by AI and one made by a real person is already a difficult task.
The Risk of Artificial Intelligence
Even though Artificial Intelligences already have great capabilities, they "still" haven't developed their own thinking, nor even critical thinking. While there was a lot of hype surrounding a recent artificial network for AI bots to communicate with each other, some videos already show that the interactions were controlled by users and relayed to the AI agents, so it wasn't something that originated from them. But does it need to reach that level for people to start considering the risks of this technology?
With the evolution of technology raising computational levels thanks to the large amount of money flowing into this sector, each month seems to be equivalent to years of development, with new technologies and hardware advancements constantly emerging from Nvidia and software companies improving their AI models. This high liquidity has made access to the latest models more democratic, and their power has increased along with the rise in computing power. Therefore, while you can use AI today to generate a cute photo, you could also use it to do something bad... Does it really need to reach the level of a conscious Skynet, or could we already have some kind of Skynet today, we just haven't realized it yet? Individual power has grown significantly since each access to AI is a processing task performed by a near-expert in the field. Wasn't the Turing Test our first warning sign?
