The headlines were troubling, to say the least.
Recently, the so-called Godfather of AI, Geoffrey Hinton, resigned from his post at Google, warning that rapid advances in artificial intelligence could ultimately threaten the survival of humanity.
“I think it’s conceivable that this kind of advanced intelligence could take over us,” said famous British-Canadian computer scientist told the CBC as it happens.
“It would mean the end of people.”
While it’s impossible to ignore such harsh comments, some experts say they risk masking more pressing and practical concerns about Canada.
“Whether intentionally or not, those who talk about the risks involved in the existence of AI, even the negative ones, are building and hyping the field,” said Western University in London. said Luke Stark, assistant professor of information and media studies. , ont.
“Given the many concerns about how these systems are currently being used by institutions, businesses and governments around the world and in Canada, I think this is a bit of a herring.”
Stark, who studies the social impact of technologies such as artificial intelligence, open letter Critical of Federal Bill C27 on Artificial Intelligence.
The letter notes that the government’s Artificial Intelligence and Data Act (AIDA), which is part of C27, lacks detail and many key aspects of the regulation on AI will need to be decided after the law is passed. I claim there is.
Renowned AI pioneer Geoffrey Hinton has resigned from Google so he can speak more freely about the potential dangers of advanced artificial intelligence. He is the latest major technologist to call for a slowdown in AI research, warning that AI could surpass human intelligence.
Experts say they look to the EU for guidance
The bill, introduced in June last year, recently completed Second reading in the House and sent to a committee for investigation.
A spokeswoman for Innovation, Science and Economic Development Canada said in a statement, “The government hopes that amendments will be proposed in response to testimony from the panel’s experts, and will make amendments to improve the bill. We are ready to consider it,” he said.
Other jurisdictions, including the European Union and the UK, are moving more quickly to introduce strong rules governing AI, experts say.
They list a long list of human rights and privacy concerns related to technology. law enforcement, misinformation and instance reinforce patterns of racism and discrimination.
The proposed law does not adequately address such concerns, said Marcia Levesque, a Harvard jurisprudence candidate who previously headed the AI and Human Rights File at Global Affairs Canada. says.
Levesque described the law as an “empty shell”. recent essaystated that it lacks “basic legal clarity”.
In an interview on Zoom, Lévesque presented a draft law covered in a blue sticky tab. Each tab shows an example of a legal clause left undefined.
“This bill leaves very important concepts to be defined in later regulation,” she said.
The bill also proposes the creation of a new commissioner to oversee AI and data in Canada.
However, Levesque said the title was a “misnomer” because unlike other commissioners, the AI and Data appointees are not independent agents, but head the regulatory body.
“From a structural standpoint, it’s really a problem,” she said.
“We incorporate protection into our innovation-driven mission, but sometimes they come into conflict.
Levesque said the EU has a “much more robust scheme” when it comes to legislation on artificial intelligence.
The European Commission will start drafting the bill in 2021 and is nearing the final stages.
Under the law, companies deploying generative AI tools such as ChatGPT are required to disclose copyrighted material used to develop the system.
Lévesque likened their approach to the checks a new plane or drug needs before it hits the market.
In Stark’s view, liberal governments are emphasizing AI as an engine of economic growth and seeking to brand Canada as an “ethical AI center.”
“To deliver on the promise of such a message, I want governments to be broader, more consultative, more involved, outside tech communities like Montreal and Toronto. ‘ said.
“quickly, slowly”
The Canadian Civil Liberties Association is one group that hopes to be heard at this upcoming consultation.
CCLA privacy attorney Tashi Alford-Duguid said, “We have not been able to obtain sufficient input from key stakeholders, minority groups, and those who we believe are likely to be unduly affected by this bill. bottom.
Fake AI-generated videos and images are becoming available online, but they are becoming increasingly difficult to find. Futurist girlfriend Sinead Bovell takes CBC’s David Common to introduce a few duped figures and show you how to spot them.
Alford-Duguid said the government needed to take a “hurry slow down” approach.
“The UK has much broader consultations. We know the EU is in the middle of very broad consultations. None of these laws seem perfect, but the Canadian government is “We are participating in this late hour. Instead, they are trying to give us such hasty and ineffective laws,” he said.
“We just look around and see that we can already do better than this.”
