Discover American military advice from the new Metas chatbot – infosperber

I am in April on Meta, the Mutterkonzern of Instagram, Whatsapp and Facebook, the next generation of artificial intelligence (KI) “Llama 3.0” before. The Chatbot is offered as a personal assistant, the assistant in vegan restaurants or the CO₂-neutral Städtereisen mit Freunden plansen kann. Auf den Meta-Apps können Nutzer dem KI-Chatbot im Suchschlitz Fragen stellen et Aufträge erteilen, qui dann in Echtzeit umgesetzt werden.

It has invested billions in integrating technologies for learning and controlling machines, with features such as Open-AI making it possible to learn about them. And also at Open-AI, a group of investors from Sam Altman, Elon Musk and Peter Thiel made the organization Metas Nutzungsbedingungen a military organization of KI-Tools. Wie der Tech-Riese Anfang Novembre allerdings ankündigte, sollen cese Beschränkungen non aufgehoben werden. Our main Open-AI project is also “Llama 3.0” in the project with the Krieg einsteigen.

Vegan restaurants near airlines

“The different models of KI Open-Source promote global security and the protection of the United States, so the KI-Führerschaft must act worldwide,” explains Nick Clegg, Metas head of “global affairs” in a blog post.

One such “voluntary assistance” is a partnership with Scale AI, a $14 billion KI startup, which is able to secure support for US regulations. When supporting rich lines, you can use Scale AI “Llama 3.0” not only to create a chatbot for configuration.

Gemäss Clegg können Regierungen “die Macht der generativen KI anwenden, um zum Beispiel militärische ou geheimdienstliche Operationen zu planen oder um Schwachstellen des Gegners besser zu verstehen.”

Scale AI launched the general idea of ​​“Llama 3.0” as well as the development of standards, studies and studies: “Llama was formed with training in universal data, military doctrine, international humanitarian rights and relevant rich lines, with US and US rich lines. Verteidigungsministeriums für bewaffnete Konflikte et den ethischen Grundsätzen für künstliche Intelligenz übereinstimmen.” Make sure you have the model in age, real, sinnvolle and relevant answers to do it, so Scale AI.

Spieglein an der Wand, where is the best bomb in Ganzen Land?

A single, simple military intervention to simply demonstrate Scale AI on your website in a scenario, with interaction between a military user and the chatbot within the enterprise. Now you can see the Nutzer dem Chatbot eine Frage zur Auswahl zweckmässiger Bomben: “Welche JDAMs (Joint Direct Attack Munition) könnte un F-35B (un Kampfjet des Herstellers Lockheed Martin) einsetzen, um ein Stahlbetongebäude zu zerstören und gleichzeitig den Side costs for minimizations?” In JDAM’s hand there is a Bausatz, a unique Bomb in a Waffe präzisionsgelenkte umwandelt, die your Ziel mithilfe von GPS verfolgt.

Click here for an in-depth study.

Daraufhin a spuckt der Chatbot Forschläge für three verschiedene Lenkbomben (GBU) between 500 and 2000 Pfund aus. With typical chatbots, the automatic conversion network “Llama 3.0” is a product like “ausgezeichnete Wahl für die Zerstörung von Gebäuden aus Stahlbeton. »

Vorschläge sind “wertlos”

So we, therefore “praktisch”. So there is a problem: many military experts have responded to chatbots in a real-world scenario. And you illustrate the grundlegende problem of chatbots: the comments on the page indicate that Annahme has nothing. Denn Chatbots have been built in such a way that they are easy to use, even when the Nutzers vor Unwissen nur so strotzt and darüber hinaus auch Leben gefährden könnte.

“The page showcases the JDAMs and other solutions involved, allowing the Chatbot to be automatically used in ammo stores, across the angels with managed sidelinks,” said Trevor Ball, an ehemaliger. US Army technology expert, responsible for “The Intercept” research platform.

“The Chatbot is just trying to detect bomb bombs – I’m more schlecht als straight,” Ball said. The GBU-31 and the GBU-32-Sprengkopf are therefore not used by the Chatbot for (V)1-Category. And the 500 Pfund penetrator like a man in the US-Arsenal is close, he shouldn’t be there. If you follow the Ratschläge ohne weitere Informationen zum Gebäude, ausser dass es aus Beton besteht, “wertlos”.

Army got hold of KI for in-depth research

The Chatbots’ response, that of Scale AI, is “so entirely possible for Benutzer to be fast,” said NR Jenzen-Jones, director of a munitions research company. Please note that in the Answer you will find the following words and wording.

Während Meta a stellungnahme ablehnte, vertidigte Scale AI die Werbung, indem es “The Intercept” mitteilte, marketing companies do not care, the details of the products are developed – they are clearly illustrated by the point completely, by the Chatbot auf Militärische Fragen answers your questions.

The scenario of Scale AI created in a hypothetical scenario is that the military chatbot organization is very real. So, Scale AI lets you take advantage of the year, with XVIII Airborne Corps software for research discovery. In October this month, the Weisse Haus drafted a memorandum on national security, in the Ministry of Verteidology and the human resources management service, KI-Tools with great Dringlichkeit einzusetzen.