😧AI Threatens Holocaust Memory

Former OpenAI chief scientist Ilya Sutskever launches new AI startup focused on safety

Key Points 

• Safe Superintelligence Inc. (SSI), is a startup dedicated to developing a safe and powerful AI system.

• Sutskever co-founded SSI with former Y Combinator partner Daniel Gross and ex-OpenAI engineer Daniel Levy.

• Unlike OpenAI, which began as a non-profit in 2015 and later restructured due to high computing costs, SSI is being established as a for-profit entity from the start.

👨🏻‍🔬News - Ilya Sutskever, the co-founder and former chief scientist of OpenAI, has founded a new AI company focused on safety. 

On Wednesday, Sutskever announced the launch of Safe Superintelligence Inc. (SSI), a startup dedicated to developing a safe and powerful AI system. He co-founded SSI with former Y Combinator partner Daniel Gross and ex-OpenAI engineer Daniel Levy.

The announcement describes SSI as a startup that advances AI capabilities and safety simultaneously, allowing rapid progress while still prioritizing safety. It notes that, unlike AI teams at OpenAI, Google, and Microsoft, SSI’s singular focus avoids distractions from management and product cycles. Furthermore, it mentions that the business model ensures safety, security, and progress are protected from short-term commercial pressures, enabling the company to scale steadily and without interruption.

🤓What's more? In an interview with Bloomberg, Sutskever provided more details about the new company but did not comment on its funding or valuation. However, unlike OpenAI, which began as a non-profit in 2015 and later restructured due to high computing costs, SSI is being established as a for-profit entity from the start.

France to build Europe’s most powerful classified supercomputer for defense AI

Key Points 

• France aims to be the best-prepared European power in military AI.

• In March, France announced plans to reallocate €2 billion from the 2024-2030 defense budget to artificial intelligence.

• Armed Forces Minister Sébastien Lecornu highlighted that, because of the sensitive nature of their work, running AI on classified material requires a completely secure network.

• This supercomputer, expected to be operational by 2025, will also be accessible to the Higher Education Ministry, other government departments, and French defense companies for running AI solutions securely. 

👨🏻‍💻News - France is set to construct Europe's most powerful classified supercomputer to enhance its AI capabilities for defense, according to Armed Forces Minister Sébastien Lecornu.

This supercomputer, expected to be operational by 2025, will also be accessible to the Higher Education Ministry, other government departments, and French defense companies for running AI solutions securely. 

🪖Looking ahead - In March, France announced plans to reallocate €2 billion from the 2024-2030 defense budget to artificial intelligence. Lecornu emphasized that, by allocating the most resources to its development, France aims to be the best-prepared European power in military AI. That being said, he highlighted that, because of the sensitive nature of their work, running AI on classified material requires a completely secure network.

Lecornu stated that AI technology is already widely used in the French military, with the Caesar howitzer using AI for drone-assisted target acquisition and the Air Force using AI in pilot training. He mentioned that going forward, France's operational army will focus on developing combat AI rather than theoretical applications, significantly changing military dynamics.

👀See also - It's worth noting that France is not the only country prioritizing AI for military applications. According to analysis by the Brookings Institution published in March, spending by the U.S. Department of Defense on AI nearly tripled to $557 million in the one-year period ending August 2023.

UNESCO: Generative AI poses threat to Holocaust history

Key Points 

• The report cautions that Generative AI could be used by malicious actors to propagate disinformation and hate-driven narratives. 

• The report pointed out instances where hackers manipulated chatbots to spread Nazi ideology.

• Generative AI models have been observed inventing events, personalities, and historical scenarios in cases where they lack sufficient data.

• UNESCO called on tech companies to adopt ethical guidelines in AI development and usage to combat misinformation and prevent misuse.

☕News - A new UNESCO report underscores the necessity of integrating ethical principles into AI development. It warns that without such measures, AI could distort the historical record of the Holocaust and contribute to the rise of antisemitism. 

The report cautions that Generative AI could be used by malicious actors to propagate disinformation and hate-driven narratives, potentially generating false or misleading content related to the Holocaust.

📒Noteworthy discoveries - 

• The report highlighted that AI has been used to distort Holocaust-related content by creating fabricated testimonies and altering historical records.

• The agency also warned about the emergence of deepfakes that deny or exaggerate the Holocaust.

• Additionally, the report pointed out instances where hackers manipulated chatbots to spread Nazi ideology. One application even allows users to interact with historical Nazi figures like Adolf Hitler.

• Generative AI models have been observed inventing events, personalities, and historical scenarios in cases where they lack sufficient data.

• According to the report, Google's Bard and ChatGPT have generated content describing Holocaust-related events that never occurred. For example, ChatGPT falsely suggested a 'Holocaust by drowning' campaign where Nazis purportedly drowned Jews in rivers and lakes. Bard created fabricated quotes from witnesses to support fictitious narratives about massacres.

🥸In conclusion - The UNESCO report called on tech companies to adopt ethical guidelines in AI development and usage to combat misinformation and prevent its misuse by malicious actors to incite violence or spread falsehoods about the Holocaust.