Deeproute.ai just announced plans to open a German operations hub in 2024, aiming to pitch its production-ready autonomous driving tech to European carmakers. Its Drive 3.0 system ditches HD maps and caps hardware costs at around $2,000 using Nvidia chips. Like many AV startups, it is pivoting from robotaxis to chase immediate cash flow from carmaker partnerships.
Intel CEO Pat Gelsinger says AI won't stay locked in cloud data centers for long, arguing that economics, latency, and privacy will push workloads to the edge. The company is set to launch its Meteor Lake PC processors with dedicated AI silicon, leveraging its 80% x86 PC market share to convince developers to run AI locally rather than pay for costly cloud compute.
Microsoft just patented a smart backpack packed with cameras, microphones, and an AI engine. The bag is designed to scan your surroundings and answer contextual queries, like checking whether a ski path is safe or adding events from a concert poster. Like most patents, it may never turn into an actual consumer product.
Elon Musk just took Tesla's unreleased FSD Beta V12 on a 45-minute live drive through Palo Alto's rush-hour traffic. Powered by pure neural networks rather than explicit code, the system handled unprotected turns and parked itself with barely an intervention. Why it matters: Musk confirmed the vision-only model runs locally on existing HW3 chips without an internet connection.
Qualcomm just revealed it is embedding generative AI into its next-gen premium mobile silicon, set to debut in late October. Processing prompts locally keeps sensitive personal data private and avoids expensive cloud server bills. Why it matters: your next premium Android phone could soon run multi-billion-parameter models completely offline without draining the battery.
Microsoft is reportedly testing AI features across built-in Windows 11 apps, including Paint, Photos, and Snipping Tool. The experiments include generative AI canvas creation and background removal in Paint, alongside OCR text extraction for screenshots and camera photos. Some features may require dedicated NPU hardware, though these integrations are still in early stages.
NVIDIA just introduced DLSS 3.5 at Gamescom, featuring a new technology called Ray Reconstruction. Instead of relying on hand-tuned denoisers, it uses an AI network trained on an NVIDIA supercomputer with 5x more data than DLSS 3 to enhance reflections and global illumination across all RTX GPUs. The upgrade will debut this fall in titles like Cyberpunk 2077: Phantom Liberty.
Chinese satellite maker STAR.VISION just launched WonderJourney-1A, an orbital satellite powered by an onboard platform called String Edge AI. Instead of beaming massive raw files back to ground stations, the satellite acts as its own brain to process observations on the fly. Why it matters: In-orbit processing could shave crucial hours off disaster relief mapping and environmental alerts.
Cruise just hit another roadblock in San Francisco after one of its driverless robotaxis collided with an active fire truck late Thursday, sending an injured passenger to the hospital. Cruise claims the car entered the intersection on a green light, though the fire department was responding to an emergency call. Why it matters: The crash comes less than a week after regulators approved 24/7 paid robotaxi rides across the city, escalating a brewing feud with local emergency officials.
NVIDIA just unveiled AI Workbench, a unified developer toolkit that lets teams test and customize pretrained generative AI models locally on RTX PCs and workstations before scaling to the cloud. Alongside the tool, NVIDIA introduced AI Enterprise 4.0 with integrated NeMo support to bridge the gap between desktop experimentation and production-grade deployments.
Apple CEO Tim Cook just pushed back on concerns that the company is missing the AI boom, insisting that AI and machine learning are already "virtually embedded in every product" Apple builds. Cook revealed Apple has been researching generative AI for years, pointing to rising R&D spending as direct proof of its massive investment. Why it matters: While rivals rush out flashy chatbots, Apple is signaling that its AI play is quietly woven straight into the devices people already use every day.
Intel CEO Pat Gelsinger just told investors the company plans to build AI into every platform and product it makes. The rollout begins later this year with Meteor Lake, its first consumer chip sporting an onboard neural processor. Why it matters: As Nvidia dominates the AI cloud, Intel is betting that latency, bandwidth, and cost limits will force AI inferencing straight onto client devices and the edge.
Qualcomm just announced a partnership with Meta to run its new open-source Llama 2 model directly on phone and PC chips starting in 2024. Instead of relying entirely on massive cloud data centers, running models locally on device could dramatically slash operating costs and unlock snappier on-device virtual assistants.
Microsoft just expanded its AI partnership with Meta, becoming the preferred cloud partner as Meta makes Llama 2 commercially available. Azure developers can now deploy and fine-tune 7B, 13B, and 70B models with built-in safety tools, while Windows devs get native optimization via ONNX Runtime and DirectML to run models locally.
Zoom just rolled out Intelligent Director for Zoom Rooms, using multiple in-room cameras and AI to track up to 16 people individually. The system automatically switches video feeds when someone turns their head or stands up to ensure good eye contact. Why it matters: It solves the awkward wide-angle boardroom view, giving hybrid teams a far more equitable seat at the table.
Unity just unveiled Muse and Sentis to weave generative AI directly into 3D game workflows. Muse speeds up asset and animation creation through text prompts and chat, while Sentis runs neural networks locally on user devices without racking up cloud computing costs. Both tools are currently available in closed beta.
Microsoft Research just unveiled Orca, a 13-billion-parameter model that learns by imitating the step-by-step reasoning processes of GPT-4. Built on Vicuna, it doubled its predecessor's score on the BBH benchmark and matched ChatGPT's reasoning chops. Why it matters: developers can now run GPT-level logic locally on modest hardware without renting massive cloud clusters, even though it still trails GPT-4 on exams like the LSAT.
Abu Dhabi's Technology Innovation Institute just rolled out Falcon, a family of open-source language models under Apache 2.0, with full Hugging Face integration. Falcon-40B took the top spot on the Open LLM Leaderboard ahead of LLaMA-65B while relying on multiquery attention to drastically cut inference memory. It gives builders a truly competitive, commercially viable open alternative.
Apple just announced iOS 17, packing on-device AI upgrades into everyday iPhone interactions. Autocorrect now runs a state-of-the-art Transformer language model directly on the Neural Engine for accurate predictions and grammar fixes, alongside on-device live voicemail transcription. Why it matters: Apple is quietly weaving edge machine learning straight into your core daily habits.
A new open-source AI music generator just landed on GitHub, turning simple text prompts and lyrics into complete vocal tracks. Early tests show clean stereo sound across styles like reggae and country, though pronunciation occasionally trips up. The model is now available to download and run locally.
