Why Ramp Up AI Development While Amplifying Its Potential Hazards?
7 Notable Departures from OpenAI: Shedding Light on Safety Concerns
Many people left OpenAI after the board problem and Altman’s return. Ilya Sutskever, Jan Leike, Leopold Aschenbrenner, Pavel Izmailov, William Saunders, Daniel Kokotajlo, and Cullen O’Keefe left because of safety worries. Chris Clark, the planning director, and Sherry Lachman, the social impact director, also left recently.
Today’s result shows that Ilya Sutskever and Sam Altman’s group have big differences they cannot fix.
There was a secret process at OpenAI. Their announcement said Altman was not fully honest with the board. Some think this means OpenAI may have made AGI internally but did not tell more people quickly. To stop the technology from being used widely before checking it is safe, Ilya and others pushed the emergency stop.
People ask those who left OpenAI, “What did you see?” But no one answers.
A person online said, “OpenAI lost its mission. Their CEO seems like a used car salesman. He said AI could make adult content. They released a flirty AI girlfriend.”
People talk a lot about AI dangers. But they don’t talk enough about human culture breaking apart. The harm of AI only shows when we use its benefits widely.
AI: A Non-Harmful Entity
AI can quickly process human knowledge. But it cannot perceive or have energy like humans. AI is just a tool to make processing faster, like computers. Its “intelligence” is only for playing games with 1s and 0s. It does not have human perception or wisdom.
AI in big areas like nuclear and biotech can cause surprises because it learns and works with data.
After a summit on AI safety in the UK, concerns about potential AI harm increased in many sectors. Especially in the US, potential AI harm is magnified globally. But deep divisions in society are ignored.
Why are we developing AI so much while amplifying its potential harm? There may be a few reasons: