When ChatGPT launched last year, it did more than just bring artificial intelligence to peoples' fingertips. It also heightened the debate within the AI community about the risks, rewards and harms presented by the technology.
This conversation on AI resulted in a widely-signed letter calling for a six-month pause in the development of more powerful AI models, and later, a statement highlighting the "extinction" risk of AI. Both invited support as well as criticism, which came from some experts who found the statement to be little more than fear-mongering.
DON'T MISS: Artificial Intelligence Isn't Going to Kill Everyone (At Least Not Right Away)
Kevin Baragona, the founder and CEO of DeepAI -- the first company to explore an online text-to-image generator -- signed the letter calling for a moratorium on the development of models more powerful than GPT-4.
The letter, he said, has largely become symbolic, although the idea of a slowdown in development due to safety concerns has caught on. OpenAI's CEO Sam Altman said recently that the company is not currently training GPT-5 due to safety risks.
"We are nowhere close to it," Altman added.
'This is Not a Joke'
Back in 2016 when DeepAI first launched, AI was little more than a "curiosity. It was a challenge. Now we've made the AI work and the challenge is a little different," Baragona told The Street. "The challenge is what should we do with this technology and how can we use it safely."
"This is not a joke," he said.
There are more than a few ways AI can harm society. AI expert Gary Marcus, for one, is more concerned with propaganda, misinformation and online fraud than he is with robots conquering the world.
But one of Baragona's largest areas of concern involves pending economic impact that "threatens to disrupt every type of knowledge work."
"It doesn't matter whether you're a doctor, a lawyer, a customer support agent, journalist, software programmer," Baragona said. "People may argue AI is not good enough. And the rebuttal to that is that AI gets better. It may be only 30% good today, but with exponential technology, you can expect it to be 99% capable in a year."
More Artificial Intelligence:
- Here's the Steep, Invisible Cost Of Using AI Models Like ChatGPT
- Why ChatGPT Can't Turn Into Marvel Villain Ultron (Yet)
- Protesters Say OpenAI CEO Is Dangerously Misled When It Comes to AGI
That is his answer to the artists, musicians and other creatives who are already being displaced by AI: "They're not alone."
The disruption posed by AI, Baragona said, is coming for every sector. The only roles not currently affected are the physical ones, such as electricians or chiropractors, but there's no safety there, either, as robotics will merge with AI to become even more prominent.
"I don't think that artists are being uniquely singled out, because everyone's equally affected," Baragona said. "I think they just now have, for the foreseeable future, AI tools in their toolbox so they should be able to create better things."
Ideal AI Regulation
Baragona's aim with DeepAI is to educate people about what the technology is capable of while advocating for certain policies. But those proposed by the EU's AI Act -- which categorizes models into unacceptable risk, high risk, limited risk and minimal risk -- are, in his mind, too restrictive. The challenge regulators are currently faced with is balancing innovation with protection.
"If we can set clear guidelines against the use of deep fakes to steal face and voice of people without permission, and similar regulation against other types of misinformation would be a great start," Baragona said. "Additionally, privacy regulation would be a great thing to have in this country."
The White House's proposed AI Bill of Rights, he said, might actually work out well if it was turned into clear law. The White House's proposition breaks up potential AI regulation into five guiding principles: safe and effective systems, data privacy, discrimination protection, notice and explanation and human alternatives.
Still, the danger on all fronts revolves around the uncertainty of this technology. Altman himself has said that a "misaligned superintelligent AGI could cause grievous harm to the world."
But that hypothetical risk of a 'singularity,' which involves AI destroying the human race, "ignores the actual harms resulting from the deployment of AI systems today," the Distributed AI Research Institute (DAIR) wrote.
Other experts, like Dr. Noah Giansiracusa, think that it is simply too soon to have a strong take either way on annihilation. Giansiracusa thinks the most likely path is one with both good and bad outcomes.
"The economists really don't know what's going to happen. Anyone who says they know what happens because of this, I think is not to be trusted," Baragona said. "There's never been a technical wave like this. You should not compare this to prior revolutions."
Though it's hard to see exactly how AI will impact things, Baragona's most confident prediction is that we are barreling toward a "sort of sci-fi future."
"I think that certainly, miracles will probably happen. I think some terrifying things are also likely," he said, adding that, just like the present moment, "it'll be a mixed bag."
Despite those potential miracles, Baragona is not "completely comfortable with our trajectory."
"If I could push a button and ensure AI was not created, I'd be kind of tempted to push that button," he said. "That said, the fact of the matter is this technology is being created. I think it's a moral imperative to try to shape the world in an ethical manner. I have a much greater ability to do this by being in the AI industry than being on the outside."
And in the meantime, the technology just keeps getting more powerful.
"We're going to continue to create AI as a species; there is a silver lining that we get to be around for the creation of a very wild development on planet Earth," Baragona said. "So I almost feel grateful that we get to witness it."
Receive full access to real-time market analysis along with stock, commodities, and options trading recommendations. Sign up for Real Money Pro now.
