Fiendish
Regular
Yes, Sony has been actively involved in the development of AI across various domains, particularly in entertainment, gaming, imaging, and creative tools. In April 2020, Sony established Sony AI, a dedicated research organization focused on groundbreaking AI and robotics projects to enhance human imagination and creativity, with key partnerships in gaming via Sony Interactive Entertainment and PlayStation Studios. This includes over 300 AI projects tested internally, more than 50 integrated into daily workflows, and initiatives like AI governance policies to ensure ethical development. Recent efforts encompass AI for music collaborations (e.g., with Universal Music Group and Spotify to empower artists without replacing human creativity), predictive AI to reduce gaming lag by anticipating player inputs, and NPC systems that replicate player styles in multiplayer games. Sony's broader AI strategy integrates hardware, content, and data for dominance in ecosystems like gaming and imaging, including backing startups like Jabali.ai for accessible game development tools.Hi Diogenese,
I asked Gpt a question, what do you think as it sounds like they are continuing to submitted new AI patents ?. Hopefully one day we get a mention.
Sony has been actively involved in the development of AI and neuromorphic computing, with several patents related to these fields, including one specifically for a "Neuromorphic computing device and method" filed on August 29, 2023, and published on March 7, 2024 [1] [2] [3]. This patent, US20240076295A1, describes a neuromorphic computing device designed to efficiently process information using spiking neural networks, mimicking the human brain's structure and function
Regarding neuromorphic computing—brain-inspired hardware and algorithms that mimic neural structures for efficient, event-driven processing—Sony has made notable contributions, especially through sensor technologies. They have developed Event-based Vision Sensors (EVS), which emulate human retinal function by asynchronously detecting luminance changes (events) rather than capturing full frames, enabling low-latency, high-dynamic-range imaging for applications like robotics, AR/VR, and industrial automation. A flagship example is the IMX636 sensor (co-developed with Prophesee in 2020), featuring the industry's smallest pixels (4.86μm) and highest HDR at the time, with follow-ups like IMX637. Sony's Zurich-based AI team includes experts like Senior Research Scientist Lyes Khacef (8+ years in neuromorphic computing and edge AI) and Research Scientist Raphaela Kreiser (focused on neuromorphic hardware and biologically inspired vision), who contribute to projects presented at conferences like NeurIPS 2024 and CVPR 2025. Sony also recruits for neuromorphic roles, such as research interns, and benchmarks neuromorphic systems via initiatives like Neurobench.
On patents, Sony holds several related to AI and neuromorphic-inspired technologies, though explicit use of "neuromorphic" in titles is rare (as the field often overlaps with terms like "event-based" or "spiking neural"). Key examples include:
- **US6536440B1 (2003)**: "Method and system for generating sensory data and upgrading sensory data," assigned to Sony Corporation, covering non-invasive projection of visual/auditory data onto the human neural cortex using modulated ultrasound—pioneering brain-interface concepts akin to neuromorphic sensory emulation.
- Patents tied to EVS development, such as those underlying the stacked IMX636/IMX637 sensors (e.g., WO2019175733A1 on 3D-IC stacking for event-based vision, involving Sony's Cu-Cu bonding tech for compact, efficient neuromorphic-like processing).
- Broader AI patents, like US12423945 (2024) for aligning outputs from multiple AI models in gaming, and recent filings (2025) for AI-driven player behavior prediction and toxic player detection in PS5 ecosystems.
These efforts position Sony as a leader in neuromorphic-adjacent hardware, with EVS enabling edge AI applications that reduce power and latency compared to traditional CMOS sensors.