Artificial Intelligence
Hassabis cuts AGI estimate to two to three years
Demis Hassabis said artificial general intelligence could arrive within two to three years, after previously putting the horizon at three to five years earlier this year, and defined it narrowly.
- Demis Hassabis said artificial general intelligence could arrive within two to three years.
- Earlier this year, he had said the timeframe was three to five years.
- He defined AGI as a system that matched the full cognitive range of the human brain.
- He said he had focused on Isomorphic Labs after stepping down from Google DeepMind and had proposed an industry-funded oversight board under government control.
Demis Hassabis said he now thought AGI could arrive within two to three years. Earlier this year, he had given a longer estimate of three to five years.
He defined AGI as a system that matched the full cognitive range of the human brain. He also said his test was whether a system trained only on pre-1901 knowledge could reproduce Einstein's 1905 papers.
Hassabis told the Wall Street Journal that his estimate kept moving closer to the present. He compared AGI's impact with an industrial revolution magnified tenfold and said it would arrive at ten times the speed.
He said he had moved on from Google DeepMind and was now concentrating on Isomorphic Labs, Google's drug discovery AI spin-off. He also said he wanted an industry-funded oversight board under government control to review models before release and warned that poor regulation would be worse than none.
Named in this story
People
- Demis Hassabis
- said AGI could arrive within two to three years
Companies
- the company that owns the drug discovery spin-off he now focuses on
- Isomorphic Labs
- Google's drug discovery AI spin-off that Hassabis now focuses on
Organisations
- Google DeepMind
- the organisation Hassabis founded and later stepped down from
How the source tells it
The piece is a short interview-led report that relays Hassabis's own expansive framing, so the register is speculative and lightly boosterish rather than alarmist.
- speculation treats an unreached AGI deadline as a near-term forecast and repeats a personal estimate as if it were a measure
- hype uses a sweeping historical comparison to magnify the expected impact of AGI