“Father of the Internet” urges restraint on AI investments

Vint Cerf, Google’s chief evangelist and the ‘Father of the Internet’, urged not to rush investments in conversational AI simply because it’s a ‘hot topic’. CNBC reports CNBC.

“There is an ethical problem here that I hope some of you will take into account. Everyone is talking about ChatGPT or Google’s version, but it doesn’t always work the way we would like,” he said.

Cerf warned against the temptation to invest solely because the technology is ‘really cool’.

“If you think: ‘Dude, I can sell this to investors because it’s a hot topic, and everyone will throw money at me’, don’t do it,” the scientist said.

He urged capital backers to be thoughtful. In his words, the problem with the technology lies in people who ‘have not changed in the last 400 years’.

“They will seek to do what benefits them, not you. So we should remember that and think about how we use technology,” Serf said, apparently alluding to human greed in general.

The scientist recounted how he once asked one of the systems to add a smiley at the end of every sentence. The AI failed the task, and Serf pointed it out. The chatbot apologized but did not change its behavior.

“We are far from awareness or self-awareness,” he said about the conversational AI.

According to the scientist, there is a gap between what the technology ‘says’ and what it does.

“This is the problem […]. You cannot distinguish a well-articulated answer from an accurate one,” says Cerf.

He gave an example of asking the chatbot to write a biography about itself. According to Cerf, the tool presented its answer as factual, though it contained inaccuracies.

From an engineering standpoint, the scientist believes developers should take responsibility and strive to reduce the harm from the technology:

“Figuring out how to minimize the potential for the worst-case scenario is very important.”

His warning came as companies such as Meta, Microsoft and Google entered the race to push chatbots to stay competitive. They have the resources to rapidly improve a technology that still frequently makes errors.

Alphabet chair John Hennessy said conversational AI systems are still far from widespread use. Chatbots face many issues with inaccuracy and ‘toxicity’ that need to be resolved before public testing, he added.

In February, OpenAI head Sam Altman called ChatGPT ‘a terrible product’.

In the same month, the founder of Microsoft said that conversational AI is the ‘main innovation’ of modern times.

Apple co-founder Steve Wozniak warned of ‘horrendous mistakes’ made by ChatGPT.