
Tech • IA • Crypto
Apple unveiled a major overhaul of Siri and strengthened its parental tools at its annual conference, against a backdrop of accelerating global investment in AI.
Fifteen years after its launch in 2011, Apple’s assistant is evolving into an AI-powered conversational model. The new capabilities aim for more natural, contextual interaction, closing the gap with recent leaders. Apple is betting on gradual integration rather than a sudden break, with performance now considered on par with everyday uses.
Apple is preparing to integrate third-party assistants like ChatGPT, Claude, or Gemini directly into Siri. A model selector appears in some versions, but the integration remains limited and less seamless than expected. The strategic challenge is to preserve its closed ecosystem while meeting demand for interoperability.
Nearly 12 minutes were devoted to child safety tools, an unusual share. New features include age-restricted accounts, parental approval of visited sites, and contact controls. Apple aims to reassure amid concerns about smartphone effects on young users while strengthening its family market.
The spatial reframing feature automatically corrects an image’s angle after capture. Less flashy than image generation, this approach aligns with Apple’s philosophy: improve what exists without distorting the experience. It nevertheless revives the debate about the boundary between real photos and recomposed images.
The growing integration of AI in photography (cleanup, upscaling, recomposition) fuels criticism. Some fear altered memories and greater confusion between real and synthetic. The question of labeling generated or modified images is becoming central for platforms.
For its Apple Foundation Models (AFM3), Apple is expanding its private cloud to Google Cloud infrastructure equipped with Nvidia GPUs. A notable shift for a historically integrated company, with implications for privacy and environmental footprint.
Beijing is preparing a massive investment vehicle in AI and tech infrastructure, estimated at $295 billion. This places the country alongside U.S. giants and intensifies global competition in compute capacity and advanced models.
The arrival of mega IPOs highlights the limits of current systems. Share delivery delays can reach 9 days, with price gaps of 10%. New tech players are trying to streamline these processes, notably via API systems and real-time infrastructure.
Between technological catch-up, tighter usage control, and geopolitical rivalries, the AI ecosystem is entering a mature phase where performance, trust, and regulation are inseparable.