Marc Andreessen pushes again on fears that AI will kill us all: ‘That is not the way it works’

• Marc Andreessen mentioned on the podcast Huberman Lab that fears round AI are overblown.
• The billionaire enterprise capitalist mentioned that AI will not “resolve to kill us all” and exchange jobs.

For some, the rise of synthetic intelligence might spark visions of a post-apocalyptic future — however one billionaire investor thinks these fears round AI are overblown.

Marc Andreessen, founding father of venture-capital agency Andreessen Horowitz that has led multimillion greenback funding rounds for quite a few AI startups, mentioned in a podcast interview earlier this week that we have to separate the “actual issues” with AI, which he mentioned are “technologically grounded,” from the “pretend issues” that he claims aren’t “rational.”

“Numerous the science-fiction situations are simply not actual,” Andreessen mentioned on an episode of Huberman Lab, a podcast led by Andrew Huberman, a neuroscientist.

Remodel expertise with studying that works

Functionality improvement is vital for companies who need to push the envelope of innovation.Uncover how enterprise leaders are strategizing round constructing expertise capabilities and empowering worker transformation.Know Extra

Among the sci-fi situations the billionaire investor refers to as “pretend issues” embrace the notions that AI will “develop the type of company” the place it would “steal our cash, partner and children,” or that AI will someday “get up and resolve to kill us all.”

“That is not the way it works,” he mentioned in response to those situations.

Andreessen additionally mentioned that “destruction of society” considerations” like how AI might bolster the unfold of misinformation, hate speech, and deep fakes aren’t “truly an actual downside,” and that financial considerations — like how AI will take everybody’s job — will not truly occur.

The actual concern, Andreessen famous, is that AI might find yourself within the palms of “unhealthy individuals” who might do “unhealthy issues” with the expertise, akin to malicious actors utilizing it to “construct higher protection instruments.”

The way in which to forestall that from taking place, he mentioned, is to make AI techniques safer and fewer susceptible to hacking.

“It’s best to have an entire new type of safety suite wrapped round you, wrapped round your knowledge, wrapped round your cash, the place you are having AI repel assaults, disinformation, hate speech, deep fakes — all that stuff,” Andreessen mentioned.

Andreessen’s ideas on AI come as tech leaders and consultants alike debate whether or not AI has the potential to develop superintelligence and exchange jobs throughout industries within the close to future.

Nonetheless, the enterprise capitalist is not too involved with these situations.

In June, two months earlier than the podcast episode was launched, Andreessen revealed a prolonged weblog publish known as “Why AI Will Save The World” the place he makes an attempt to dispel the fears round AI’s dangers to humanity.

Within the publish, Andreessen wrote that AI can “make all the things we care about higher,” and that AI firms should have the ability to construct their applied sciences “as quick and aggressively as they will” in an effort to “maximize its features for financial productiveness and human potential.”

Whereas Sam Altman, the CEO of OpenAI, the corporate behind ChatGPT, mentioned {that a} authorities company must be shaped to supervise AI security, Andreessen disagrees. He mentioned the non-public sector must work with governments to deploy the expertise safely.

“AI might be an extremely highly effective device for fixing issues, and we must always embrace it as such,” Andreessen wrote.

Andreessen did not instantly reply to Insider’s request for remark when contacted by way of his VC-firm, A16z