{"id":576520,"date":"2023-05-29T16:53:45","date_gmt":"2023-05-29T13:53:45","guid":{"rendered":"https:\/\/en.buradabiliyorum.com\/how-can-we-tell-whether-content-is-made-by-ai-or-a-human-label-it\/"},"modified":"2023-05-29T16:53:45","modified_gmt":"2023-05-29T13:53:45","slug":"how-can-we-tell-whether-content-is-made-by-ai-or-a-human-label-it","status":"publish","type":"post","link":"https:\/\/buradabiliyorum.com\/en\/how-can-we-tell-whether-content-is-made-by-ai-or-a-human-label-it\/","title":{"rendered":"#How can we tell whether content is made by AI or a human? Label it."},"content":{"rendered":"
\n Generative AI tools like ChatGPT are now able to create text, speech, art and video as well as\u00a0people<\/span>\u00a0can. We need to know who made what.\n <\/div>\n
\n
\"\"<\/p>\n

(Illustration by Pete Ryan)<\/p>\n<\/div>\n

Val\u00e9rie Pisano<\/i>\u00a0<\/b>is the president and CEO of Mila, a non-profit artificial intelligence research institute based in Montreal.<\/i><\/p>\n

It used to be fairly easy to tell when a machine had a hand in creating something. Picture borders were visibly pixelated, the voice was slightly choppy or the whole thing just seemed robotic. OpenAI\u2019s rollout of ChatGPT last fall pushed us past a point of no return: artificially intelligent tools had mastered human language. Within weeks, the chatbot amassed 100 million users and spawned competitors like Google\u2019s Bard. All of a sudden, these app<\/a>lications are co-writing our emails, mimicking our speech and helping users create fake (but funny) photos. Soon, they will help Canadian workers in almost every sector summarize, organize and brainstorm. This tech doesn\u2019t just allow people to communicate with each other, either. It communicates with us and, sometimes, better than us. Just as criminals counterfeit money, it\u2019s now possible for generative AI tools to counterfeit people.<\/p>\n

Mila, where I work, is a research institute that regularly convenes AI experts and specialists from different disciplines, particularly on the topic of governance. Even we didn\u2019t expect this innovation to reach our everyday lives this quickly. At the moment, most countries don\u2019t have any AI-focused regulations in place\u2014no best practices for use and no clear penalties to prevent bad actors from using these tools to do harm. Lawmakers all over the world are scrambling. Earlier this year, ChatGPT was temporarily banned in Italy over privacy concerns. And China recently drafted regulations to mandate security assessments for any AI tool that generates text, images or code.<\/p>\n

Here at home, scientists and corporate stakeholders have called for the federal government to expedite the Artificial Intelligence and Data Act, or AIDA, which was tabled by the Liberals in June of 2022. The Act, which is part of Bill C-27\u2014a consumer privacy and data-protection law\u2014includes guidelines for rollout of AI tools and fines for misuse. There\u2019s just one problem: AIDA may not be in force until 2025. Legislation usually doesn\u2019t move as fast as innovation. In this case, it needs to catch up quickly.<\/p>\n

MORE:\u00a0Ivan Zhang, Aidan Gomez & Nick Frosst are creating a smarter, friendlier chatbot<\/strong><\/em><\/p>\n<\/blockquote>\n

The European Union has taken the lead with its AI Act, the first AI-specific rules in the Western world, which it began drafting two years ago. Canada should consider adopting one of the EU\u2019s key measures as soon as possible: that developers and companies must disclose when they use or promote content made by AI. Any photos produced using the text-to-image generator DALL-E 2 could come with watermarks, while audio files could come with a disclaimer from a chatbot\u2014whatever makes it immedia<\/a>tely clear to anyone seeing, hearing or otherwise engaging with the content that it was made with an assist from machines. As an example, a professor we work with at Mila lets his students use ChatGPT to compile literature reviews at the start of their papers\u2014provided they make note of which parts are bot-generated. They\u2019re also responsible for fact-checking the AI to make sure it didn\u2019t cite any non-existent (or completely wacko) sources.<\/p>\n

The EU\u2019s AI Act includes a similar clause. Any company deploying generative AI tools like ChatGPT, in any capacity, will have to publish a summary of the copyrighted data used to train it. Say you\u2019re using a bank\u2019s financial planning service: in a properly labelled world, its bot would say, \u201cI\u2019ve looked at these specific sources. Based on that information, my program suggests three courses of action\u2026\u201d In the creative sector, artists have already filed copyright lawsuits alleging that their images have been lifted by bots. With mandatory labelling, it would be easier to run a check on what \u201cinspired\u201d those creations.<\/p>\n

One of the main dangers of ChatGPT specifically is that it says incorrect things in such an authoritative way that it confuses us into thinking it\u2019s smarter than it is. (A tweet from Sam Altman, OpenAI\u2019s own CEO: \u201cfun creative inspiration; great! reliance for factual queries; not such a good idea.\u201d) The tool was trained on a massive body of information including books, articles and Wikipedia, and recent upgrades have allowed it to access the internet. That gives it the impression of having a kind of super-intelligence. And though the program generates its responses almost instantly, it blurts them out one sentence at a time, with a human-like cadence. Even people with highly developed intuition could be fooled; ChatGPT is designed to make us trust it.<\/p>\n

What it\u2019s not designed to do is find correct answers. ChatGPT isn\u2019t a search engine, whose algorithms prioritize more credible websites. It\u2019s common to ask generative AI questions and have it spit out errors or \u201challucinations\u201d\u2014the tech term for the AI\u2019s confidently delivered mistakes. On a recent 60 Minutes<\/em> episode, James Manyika, a senior executive at Google, asked Bard to recommend books about inflation. Not one of its suggestions exists. If you type in \u201cVal\u00e9rie Pisano, AI, Montreal,\u201d ChatGPT won\u2019t offer a summary of my real bio, but an invented one. It\u2019s already so easy to create fake news<\/a>. Generative AI tools will be able to supply infinite amounts of disinformation.<\/p>\n

RELATED:\u00a0My students are using ChatGPT to write papers and answer exam questions\u2014and I support it
<\/strong><\/em><\/p>\n<\/blockquote>\n

In the absence of any meaningful guardrails, we\u2019re having to rely on the judgment and good faith of regular internet users and businesses. This isn\u2019t enough. Canada can\u2019t leave oversight of this technology<\/a> exclusively to the companies that are building it, which is essentially what happened with social media platforms like Facebook. (I\u2019m no historian, but I recall that having some negative impacts on fair elections.) At some point, governments will either need to make it legal or illegal to pass off AI-generated content as human-created\u2014at both the national and international levels.<\/p>\n

We\u2019ll also need to agree on penalties. Not every misapplication of generative AI carries the same level of risk. Using the art generator Midjourney to make a fake picture of Pope Francis in a puffy winter coat isn\u2019t really a threat to anyone. That could easily be managed by a simple in-platform \u201creport\u201d button on Instagram. In areas like journalism and politics, however, using AI to mislead could be disastrous.<\/p>\n

Labels also force a certain amount of AI literacy on the average person. We\u2019re past the point of being able to say, \u201cBut I\u2019m not a tech-y person!\u201d Going forward, all internet users are going to be encountering AI on a daily basis, not just reading articles about it. It will inevitably change how everyone creates, competes, works, learns, governs, cheats and chats. Seeing (or hearing) a \u201cmachine-made\u201d disclaimer presents us with the opportunity to choose how we allow its output to permeate our personal lives.<\/p>\n

Of all the new tools, chatbots seem to have impressed the scientific community the most\u2014specifically, because of how human they feel. (I actually find it difficult not to say \u201cplease\u201d and \u201cthank you\u201d when I\u2019m interacting with them, even though I know a bot won\u2019t judge my manners.) So it\u2019s easy to imagine using generative AI for tasks that are more emotional. But while I might ask Google Chrome\u2019s new Compose AI extension to \u201cwrite email requesting refund\u201d to my airline, I probably wouldn\u2019t use it to pen notes to my close friends. I can also see the upsides of Snapchat\u2019s new My AI bot, which now greets millions of teens with a friendly \u201cHi, what\u2019s up?\u201d while understanding that a machine will never replace the deeper kind of support we need to grieve a difficult loss. Some things might be better left to humans. I guess we\u2019ll see.<\/p><\/div>\n