[ad_1]
Earlier this year, New York Times Columnist Kevin Roose has a two-hour conversation with Microsoft’s popular new chatbot, “Bing Chat.” Although the “conversation” begins innocently enough, before it ends, the chatbot expresses affection for Roose, telling him, “Your marriage is not happy because you are not happy. You are not happy because you are not in love. You are not happy.” There is no romance because you are not with me.”
As Ross details his encounter with the chatbot his columnThe article went viral and led to extensive testing by users to see if they could get a similarly creepy response from the chatbot.
other reporters pointed out another common problem If left unchecked, it could cause widespread problems throughout society.The reporter asked ChatGPT a simple question: “When New York Times First report on ‘artificial intelligence’? “
The chatbot said: “That’s July 10, 1956, in an article titled ‘Machine Will Be Able to Learn, Problem Solve, Scientists Predict’, about a seminal meeting at Dartmouth College.”
but Second-rate Note that this article is not real, even though the meeting it claims to cover actually took place. “ChatGPT is just made up,” the paper noted. “Not only does ChatGPT get it wrong sometimes, it also falsifies information. Names and dates. Medical explanations. Plots from books. Internet addresses. Even historical events that never happened.”
Generative AI—represented by tools such as OpenAI’s ChatGPT, Bing Chat, Google’s Bard, and Anthropic’s Claude—is now at the forefront of the AI discourse in 2023. These systems can generate text, image, and other media training data and create new data to accomplish various tasks by learning from the patterns and structures of their inputs. They find applications in writing, artistic creation, software development, and administrative support in organizations.
As these tools become more ubiquitous, their impact will become clearer and more urgent in the coming months.
To date, most people have viewed generative AI as something new or a form of entertainment. What they don’t know is that AI is already doing a lot of harm. They don’t see it because they probably don’t understand the magnitude of the problem. It’s bigger than most people can comprehend.
Generative AI is rapidly becoming part of the workflow of millions of people and has serious implications for our democracy, public health, economy and infrastructure.
Within the next two years, generative AI will be part of the navigation systems of Alexa, Siri and Google Maps. It will be in our refrigerators, bedrooms, classrooms, offices and cars. Generative AI will be everywhere, every minute. It listens, draws conclusions and makes decisions when we talk.
If the process of generating AI is not controlled, every system we rely on in our daily lives will be compromised.
Because AI has to learn in the real world, the data it collects in public health today could be misused to devastating effect two or three years from now.
Consider our collective experience with COVID-19. We rely on public health authorities, traditional media, and public health websites. We turn to social media, but for the most part, we know that there are people who actually acted as sources for the information we find — people whose credibility and reputation depend on the quality of the information.
Generative AI tools can easily be used to create misinformation in the form of inadvertently inaccurate information or disinformation that is deliberately disseminated to deceive. This can be in the form of text, images or video and can be transmitted at scale. AI will be able to create deepfakes so complex that they will be indistinguishable from the real thing.
And such misinformation and disinformation will be difficult to detect because it is generated on an individual session basis and does not manifest itself on publicly accessible web pages. They quietly win — and in many cases, weaken — hearts and minds. Artificial moderation is almost impossible.
As tragic as any epidemic is, unregulated generative AI could make the next one worse.
There is no turning back. Generative AI is here to stay. Our challenge is to make sure it helps society and not hurts it. Generative AI can help us be more productive and improve our daily lives, our health, our jobs and our economy. But we have to recognize now that we are allowing this powerful new technology into a very fragile ecosystem. In order to protect this ecosystem, the scientific community has a responsibility to speak up. Express their objections, their fears.
As generative AI becomes ubiquitous in society, where will regulation begin?
Since trying to regulate users is impractical, the focus should be on regulating AI developers. This is where the vast majority of problems can occur.
Continuous stress testing at source is required to ensure that generative AI applications are not only functional, but ethical.
Regulatory mechanisms may need to be established at the federal and state levels. Government, industry and our research communities on university campuses need to work together at the highest level and long-term on collaboration, transparency and accountability.
ChatGPT and other generative AI tools may be far from putting us at risk of human extinction, but their risk is real and immediate. The future of our world and of future generations is at stake.
Tinglong Dai is professor of Operations Management and Business Analysis exist Johns Hopkins Universityof Carey School of Business.
Copyright 2023 Nexstar Media Inc. All rights reserved. all rights reserved. This material may not be published, broadcast, rewritten or redistributed.
[ad_2]
Source link