Neanderthals and AI: Myths and Technological Realities

⚡

Key Takeaways

1Geneticists are questioning the theory of interbreeding between Homo sapiens and Neanderthals, proposing an explanation based on population structure.

2The growing role of AI in military conflicts raises doubts about the effectiveness of human oversight, as the Pentagon and Anthropic clash over this issue.

3The White House is seeking access to Anthropic's controversial Mythos model, despite national security concerns.

💡Why it matters — These debates illustrate the tensions between scientific and technological advancements and the ethical and security implications they entail.

The Questioning of Our Neanderthal Heritage

The idea that we all carry an "inner Neanderthal" has long fascinated both the public and scientists. This theory is based on the hypothesis that Homo sapiens and Neanderthals interbred at some point in history, leaving a genetic imprint in our modern DNA. This discovery has been hailed as one of the most significant of the 21st century in the field of human evolution. However, in 2024, geneticists proposed a new perspective that could upend this well-established theory. They suggest that what is considered evidence of interbreeding could actually be explained by population structure, meaning how genes are distributed and concentrated in isolated groups. This hypothesis challenges the very foundations of our understanding of Neanderthal heritage.

The Illusion of Human Oversight in AI Wars

Artificial intelligence is increasingly integrated into modern military strategies, raising crucial questions about the role of humans in these systems. AI is at the center of a legal battle between Anthropic and the Pentagon, playing an increasing role in the conflict with Iran and raising questions about the extent to which humans should remain "in the loop." According to Pentagon guidelines, human oversight is supposed to provide accountability, context, and security. But the idea of "humans in the loop" is a comforting distraction. The real danger is not that machines act without supervision; it’s that human supervisors have no idea what the machines are actually "thinking." Fortunately, science may offer a way forward.

Current Technological Stakes

The White House and the Mythos Model

Despite the blacklisting of Anthropic, the White House is closely interested in Mythos, an AI model deemed too dangerous for public release. Trump officials are negotiating access to Mythos. Anthropic has stated that it is too dangerous for public publication. Finance ministers are alarmed by security risks. Anthropic has just launched a less risky model than Mythos. The Pentagon has waged a cultural war against the company.

Sam Altman and Conflicts of Interest

The side activities of Sam Altman, a leading figure at OpenAI, raise concerns about potential conflicts of interest. His personal investments could influence OpenAI's strategic decisions. A jury will soon decide whether OpenAI has abandoned its founding mission. The company is making significant strides in science.

Pentagon's Dependence on SpaceX

A recent Starlink outage during drone testing highlighted the Pentagon's dependence on SpaceX. This incident is just one example among others of the disruptions faced by the Navy. The DoD is also turning to Ford and GM for military innovations.

Delays in AI Expansion

Delays in building data centers threaten to slow down AI expansion, with 40% of this year's projects at risk of falling behind schedule. This issue is partly due to local communities' reluctance to welcome these infrastructures in their neighborhoods.

Advances from Alibaba and Google

Alibaba recently unveiled Happy Oyster, a global model aimed at improving AI's understanding of physical reality. However, challenges remain, particularly in understanding cause-and-effect relationships. Meanwhile, Google's Gemini now generates AI images tailored to personal data by analyzing users' Google services and data.

OpenAI and Agency

OpenAI has strengthened its coding system with an update to Codex, directly targeting Claude Code. While this advancement is promising, it does not convince all experts in the AI coding field.

European and Cultural Initiatives

Europe has launched a free online age verification application for businesses, while smart glasses give Korean theaters hope for a K-Pop moment. Finally, voice actors worldwide are fighting against the AI surge in Hollywood, as their voices form the models that replace them.

Quote of the Day

"There is this dark period between now and some point in the future where the advantage is heavily skewed towards offensive AI." — Rob Joyce, former director of cybersecurity at the National Security Agency, warns against new hacking threats created by AI.

The Race for Rare Earth Elements

Access to rare earth elements is crucial for achieving emission reduction goals and non-fossil energy production. China dominates this market, while extraction in the United States remains limited, prompting scientists and companies to explore alternative sources. This quest is essential to ensure the global energy transition.

A Bit of Lightness

To brighten your day, here are a few suggestions: a ska cover of Rage Against the Machine to kickstart a musical revolution, discovering how far Stretch Armstrong can stretch, or customizing ambient sounds to soothe the mind. Finally, an inspiring story of a young girl guiding a seal to perform tricks shows that childhood dreams can come true.