ADVERTISEMENT

OpenAI Chief Sam Altman Expects Superintelligence Soon As 'Deep Learning Worked'

The OpenAI CEO suggests that superintelligence will be possible in a few years though he caveats that it may take longer.

<div class="paragraphs"><p>In his essay, Sam Altman makes allusions to prosperity, explaining that humanity built increasingly capable AI because "deep learning worked, got predictably better with scale, and we dedicated increasing resources to it." (Photo source: Flickr)</p></div>
In his essay, Sam Altman makes allusions to prosperity, explaining that humanity built increasingly capable AI because "deep learning worked, got predictably better with scale, and we dedicated increasing resources to it." (Photo source: Flickr)

OpenAI CEO Sam Altman outlined his vision for what he terms the 'Intelligence Age' in a rare blog post published on September 23. In the post, Altman argued that artificial intelligence would be able to help humanity "accomplish much more than we ever could without AI". This he claims will eventually lead to personal teams of AI which will be used to build whatever is imagined.

In his essay, the OpenAI CEO suggests that superintelligence will be possible in 'a few thousand days' though he admits that it may take longer. For context, a few thousand days is as little as three years.

Superintelligence is a concept that suggests that machine intelligence might one day surpass humanity, though how soon that will happen is hotly debated among experts in the field of AI.

Throughout his essay, Altman makes allusions to prosperity, explaining that humanity built increasingly capable AI because "deep learning worked, got predictably better with scale, and we dedicated increasing resources to it."

He admits that while there are several details still left to "figure out" he's confident that the remaining problems will be solved.

Opinion
Indian Educators To Procure 1,000 AI-Powered EdTech Innovation

Altman's blog post comes barely two weeks after OpenAI announced a new family of AI models and released a preview version of the same, titled o1-preview. These new models are being trained via reinforcement learning and 'chain of thought' in order to answer queries. OpenAI has said that the new family will also be less likely to hallucinate—make up information— while responding to queries.

More pertinently perhaps, Altman's blog post looks suspiciously well-timed. OpenAI is looking to close another funding round, one that'll bump its current valuation from around $80 billion to $150 billion.

The company is looking to raise anything between $5 billion to $7 billion, with a minimum investment of $250 million from investors, according to a report from The Information. New York-based Thrive Capital is leading the funding round, offering up $1.25 billion, according to Bloomberg.

Opinion
Artificial Intelligence May Replace Some Jobs, Employees Need Retraining: IT Leaders