A few days ago, a Weibo blogger posted a photo of himself and wanted to show off his recent new hairstyle. Unexpectedly, the comment Robert directly followed: "Do you think that changing your hairstyle can remedy your ugly appearance?" Comments. As soon as this poisonous tongue replied, it immediately attracted onlookers and heated discussions among netizens.
Some netizens expressed strong dissatisfaction and questioned whether the comments on Robert were completely out of control. Some netizens believe that commenting on Robert is just completing the positioning of the "conspicuous bag" handed over to it by Weibo, and wants to attract more attention through this cold and trembling creative method. But many netizens pointed out that this has clearly violated basic Internet etiquette.
Some analysts have also pointed out that the reason why Commenting on Robert made such radical remarks has a lot to do with its training data. In the past, on Weibo, many netizens were accustomed to commenting on others in a sharper tone, which also directly affected the results of commenting on Robert's training.
And Weibo, as the operator of the comment Robert, should post-process its language model and restrict some unfriendly content that may cause physical and psychological obstacles to others? This has also become an important issue for netizens to discuss extensively. What should be the constraints on the interaction between AI and humans on the Internet?
As a machine intelligent assistant, should greater freedom of expression be given in the absence of clear malice? Or is there a need for clearer rules for self-censorship? There are also some opinions that any open system will inevitably face various marginal case tests in the early stages of operation.
For commenting on Robert, after this controversy, it will have a deeper understanding of the sensitive issues of online expression, and make corresponding adjustments in the follow-up service, and the final service level will be greatly improved. Some netizens joked that commenting on Robert might as well use it as an offline "complaint box".
only opened the Weibo private message function for it, and let netizens write a long story about it in the private message, pouring out their dissatisfaction and anger. It's more direct than a public comment. Over time, it seems that Robert is also constantly learning and refining himself. Recently, the remarks it replies have become more and more pertinent and rational, and it is rare to see those yin and yang expressions anymore.
Some netizens described it as if it had really grown up and knew what to say and what not to say. In any case, to a certain extent, this turmoil has prompted people to think more deeply about the ability of AI to express itself openly in the online environment. Only on the basis of respect and understanding can the interaction between man and machine move towards a more harmonious direction.
Learning to humanize respect is also a long-term process for review Robert and other AI reply bots. It requires accumulating experience in daily interactions, distinguishing between what kind of statements hurt others and what kind of expressions can win everyone's acceptance.
It also requires continuous guidance and oversight from the organic intelligence company and the operations team to avoid some unforeseen serious problems. Netizens also need to be appropriately tolerant and understanding when interacting with AI. After all, AI is not a "human" in the true sense of the word, and its speech is not as subjective as that of humans.
As long as all parties participate in it in an open and positive manner, AI will become more and more "human-like" as technology and interactive capabilities continue to grow. At that point, humans and AI may be able to communicate with each other as freely as old friends. This will undoubtedly be an extremely important step towards building a smarter society.