Google DeepMind Scientist Argues AI Will Never Be Conscious

Alexander Lerchner, a senior staff scientist at Google DeepMind, published a paper arguing that no AI or computational system can become conscious. Titled "The Abstraction Fallacy: Why AI Can Simulate But Not Instantiate Consciousness," the paper contends that AI systems are "mapmaker-dependent," requiring human agents to organize data. Lerchner claims that artificial general intelligence (AGI) is possible without sentience, describing it as a "non-sentient tool." This contradicts statements by DeepMind CEO Demis Hassabis, who predicted AGI would have an impact "10 times the Industrial Revolution." Philosophers and researchers like Johannes Jäger and Mark Bishop note that similar arguments have existed for decades and criticize Lerchner for failing to cite prior work. The paper was originally published on March 10 and remains on Google DeepMind's site, though its PDF was later modified to remove Google branding after media inquiries. Emily Bender, a linguistics professor, suggests the paper would have faced peer-review scrutiny in a normal academic process. The debate highlights tensions between AI company narratives and rigorous philosophical analysis.