AI-generated summary
Yapay zeka teknolojileri hızla gelişirken, güvenlik ve etik kullanımı konusunda endişeler artıyor. Bill Gates, teknoloji dünyasının önde gelen figürleri arasında yapay zekanın potansiyel tehlikeleri hakkında sıkça uyarıda bulunan bir isimdir.
Microsoft’un kurucusu Bill Gates, NBC’nin Meet the Press programına verdiği röportajda yapay zekanın yol açabileceği tehlikeler hakkında konuştu. Gates, teknolojinin ''bir milyar kişinin ölümüne neden olabilecek olayları tetikleyecek kadar güçlü'' olduğunu belirtti.
Gates, tehdidi yapay zekanın kötü niyetli kişiler tarafından kullanılmasına bağlayarak, ''Kötü niyetli insanların en yeni yapay zeka araçlarıyla birleşimi kadar güçlü bir silah hiç olmadı.'' dedi.
''ŞİRKETLERİN KENDİLERİNİ DENETLEMESİ YETMEZ''
Yapay zeka şirketlerinin kendi güvenlik kurallarını belirlemesinin yeterli olmadığını söyleyen Gates, “Kimse öz denetimin yeterli olduğunu düşünmüyor.” ifadesini kullandı.
Gates’e göre güvenlik önlemleri ve denetimin nasıl yapılacağına ilişkin tartışmaya siyasetçiler ile kolluk kuvvetleri de katılmalı ve yapay zekâ yasayla düzenlenmeli.
AI outlook — possibilities, not facts
Yapay zeka güvenliği konusunda uluslararası düzenleme tartışmaları başlayacak.
Likely · Within months

The joint statement prepared by the Netherlands' AIVD, Military Intelligence and Security Service, NCSC, NCTV, Public Prosecutor's Office, CIO Rijk and the National Police Service stated that artificial intelligence accelerates and strengthens cyber attacks, causes long-term disruptions in vital systems in cases where defenses are inadequate, and disinformation has become personalized. It was emphasized that attackers can detect and exploit vulnerabilities faster by using existing and easily accessible artificial intelligence models, and that defense should be cultural and awareness-oriented rather than technical measures.

PrismML developed a 1-bit Bonsai LLM model with 2 billion parameters optimized for smart glasses running on Qualcomm's Snapdragon platform. The model increases data privacy by running locally on the device and aims to deliver efficient AI processing despite hardware limitations. There is currently no commercial smart glasses model on the market that includes this software.

Apple's new flagship iPhone 18 Pro Max was compared to the iPhone 17 Pro with its A20 Pro processor, smaller Dynamic Island, Face ID moved under the screen and a 48 MP main camera with variable aperture. The performance increase is between 22%-26%, the battery capacity has been increased to 5391 mAh and 60W fast charging support has been added. Although it is difficult to notice a difference in speed in daily use, a serious presentation is provided for professional video editing and native artificial intelligence users.

The new generation model of Apple TV 4K will come with the A17 Pro processor, 8GB RAM and Wi-Fi 7 support. The image leaked by MacRumors confirmed the existence of the device. The gaming and Siri experience will be strengthened with hardware improvements. The price and official launch date are still uncertain.

The Steam review bombardment, which started after a player was penalized for his Atatürk profile photo in the Hearts of Iron IV game, was removed from the overall score calculation with the intervention of Valve.

Microsoft has decided not to use the Copilot Plus PC brand, which it uses for artificial intelligence-supported computers, starting from the new Surface models. The company is preparing to announce its new strategy on October 7.