AI Chatbots: The Good, the Bad, and the Ugly

2023-05-18

DataSolutions' Francis O'Haire believes that the competition between Google's Bard and OpenAI's ChatGPT will drive advancements in artificial intelligence technology. Artificial intelligence chatbots have garnered significant media attention. When OpenAI launched ChatGPT in November 2022, it was seen by many as a major technological breakthrough—a form of artificial intelligence that can provide detailed responses and coherent answers across multiple domains of knowledge. However, like all groundbreaking technological advancements, it has both positive and negative aspects—splitting the atom brings both the hope of cheap energy and the threat of nuclear destruction. In fact, even leaders of large tech companies find themselves at a crossroads when it comes to their views on artificial intelligence. Previously, there were reports of Bill Gates criticizing Elon Musk and over a thousand tech leaders who expressed concerns that artificial intelligence could destroy humanity and called for a pause in its development. Gates believes that halting the development of systems like ChatGPT does not solve any problems—so what should industry leaders focus on and how can they leverage new technologies in this field? When it comes to chatbot technology, the "good" aspects are clearly the most important, but what about the "bad" and "ugly" aspects? Used for Crime In March of this year, the European law enforcement agency released a report showcasing how ChatGPT can be used for criminal activities. One key point of the report is that "while all the information provided by ChatGPT can be freely obtained on the internet, using the model to provide specific steps by asking contextual questions means that criminals can better understand and carry out various types of crimes." For example, phishing attempts are often limited by the criminals' ability to make the information appear authentic—you might see spelling and grammar mistakes or false branding. However, technologies like ChatGPT can greatly assist criminals in quickly and massively producing phishing emails that appear more professional, eliminating the aforementioned limitations. Furthermore, ChatGPT can generate computer code without the need for users to have a strong programming background. While there are appropriate safeguards in place to restrict the creation of malicious code, these safeguards have not yet been proven foolproof. Cleverly crafted questions can manipulate artificial intelligence to produce things it shouldn't. Check Point Research recently documented an example of cybercriminals using ChatGPT to enhance the effectiveness of existing malware. Manufacturing Fake News We often hear about "fake news." Artificial intelligence has proven itself capable of generating realistic photos and images based on user requests. These can be used to propagate false information or fake news to promote the agendas of individuals, groups, or even governments. The most typical example is the fake photo of former President Trump being violently arrested and detained by the police. Let's not forget about deepfakes, many of which are pornographic—studies show that in these manipulated photos, 99% of the faces of female celebrities are mapped onto porn stars. There is also a less provocative but attention-grabbing fake photo of Pope Francis wearing an impressive designer coat, which went viral online and deceived many people around the world. With Google's introduction of Bard to compete with ChatGPT, the chatbot field will become increasingly prosperous, which will only drive advancements in artificial intelligence technology and continue to contribute to doing the right things. After all, artificial intelligence and machine learning have been good tools for many years, and the rapid development of new tools like ChatGPT, Bard, and others will undoubtedly be put to good use. In the business realm, chatbots with artificial intelligence and machine learning capabilities can help redefine customer service on a large scale. They can handle thousands of customer queries in a short amount of time and improve average response times. They can also help businesses maintain consistency in answering questions and enhance the customer experience with the brand—they can even support multiple languages. We have already seen cybersecurity firms widely using these technologies to combat hackers and alleviate the workload of overloaded cybersecurity teams. Artificial intelligence tools like ChatGPT can be incredibly powerful in helping people deal with increasingly complex threats. After all, if bad actors find ways to exploit this powerful technology, the rest of us must do the same to create a fair competitive environment.

RECENT AI NEWS

RECENT AI TOOLS