ChatGPT, an artificial intelligence-driven chatbot (聊天机器人) programme that responds to user prompts, has dominated social networks in recent weeks. Its _________ is such that within days of its launch, OpenAI, the artificial intelligence company behind the chatbot, announced that 1 million people had logged on. In comparison, Facebook and Spotify took months to attract that level of _________. The attraction is obvious: ChatGPT can generate jokes, essays and computer code from a short writing prompt.
There’s nothing _________ in software that produces fluent and logical writing. ChatGPT’s predecessor, the Generative Pretrained Transformer 3 (GPT-3), could do that too. Both were trained on an unimaginably large amount of data to answer questions in a _________ way. But ChatGPT has been fine-tuned by being fed the data on human “conversations”, which significantly _________ the truthfulness and informativeness of its answers.
_________, ChatGPT still produces answers that seem reasonable but are incorrect or nonsensical. Even the bot itself says, “My responses are not intended to be taken as _________.” This might be a big problem on the Internet, as many web platforms _________ the capability to protect themselves against a flood of AI-generated content. Such websites as Stack Overflow, where users can find answers to programming-related questions, had to ban ChatGPT-produced posts, as its human moderators (管理员) could not deal with the _________ of believable but wrong replies. It is __________ to give out tools that could be used to mass produce fake news and messages.
Letting loose ChatGPT raises the question of whether content produced by it can be truly trusted. A human author is __________ their work in a way AI is not. Artificial intelligence is not artificial __________: ChatGPT does not know what it is doing; it is unable to say how or why it produced a response; it has no grasp of human experience, and cannot tell if it is __________. While OpenAI has safeguards to refuse inappropriate requests, such as to tell users how to commit crimes, these can be bypassed. AI’s potential for harm should not be __________. In the wrong hands, it could be a weapon of mass destruction.
No one knows what will happen to such technologies if the software engineers of the future will themselves be software programs. “Human supervision and __________ are required,” OpenAI’s researchers have warned.
【小题1】A.attempt | B.appeal | C.purpose | D.solution |
【小题2】A.engagement | B.point | C.competition | D.ambition |
【小题3】A.harmful | B.positive | C.novel | D.obvious |
【小题4】A.creative | B.general | C.simple | D.believable |
【小题5】A.converted | B.impressed | C.increased | D.limited |
【小题6】A.In brief | B.As a result | C.Even so | D.Loosely speaking |
【小题7】A.fact | B.routine | C.discrimination | D.details |
【小题8】A.expel | B.lack | C.expect | D.improve |
【小题9】A.quality | B.potential | C.volume | D.fear |
【小题10】A.experimental | B.risky | C.unnecessary | D.reliable |
【小题11】A.scared of | B.open to | C.pleased with | D.responsible for |
【小题12】A.consciousness | B.innovation | C.emotion | D.curiosity |
【小题13】A.fitting in | B.hanging on | C.making sense | D.taking advantage |
【小题14】A.rejected | B.blamed | C.underestimated | D.contributed |
【小题15】A.civilization | B.caution | C.loss | D.pressure |