NYC Mayor Eric Adams just revealed his office has been using AI to clone his voice for automated robocalls in languages he doesn't actually speak, including Spanish, Mandarin, and Cantonese. Critics are slamming the undisclosed deepfakes as deeply unethical and Orwellian, though Adams dismissed the pushback, arguing it's just a practical way to reach non-English-speaking residents.
Baidu CEO Robin Li just unveiled Ernie 4.0 at an event in Beijing, boldly claiming the new model is not inferior to OpenAI's GPT-4 in any aspect. The update brings major upgrades to understanding, reasoning, generation, and memory, and will plug directly into Baidu Search and Maps. Markets reacted cautiously, with shares dipping slightly after the demo.
Dubai Police just revealed plans to deploy autonomous electric patrol rovers across residential areas. Built by Micropolis Robotics, the vehicle features 360-degree facial recognition, an onboard drone, and AI trained to detect suspicious behavior across a 15-hour shift. Don't expect any high-speed chases, though—its top speed tops out at a leisurely 4.35 mph.
Anthropic just expanded web access to claude.ai, officially rolling out its flagship AI assistant to users across 95 countries for professional and everyday tasks. It’s a major leap forward from its initial US- and UK-only availability, giving millions more users direct access to Claude.
Inflection AI just gave its conversational assistant Pi real-time web access. Users can now ask Pi about breaking news and live events directly. The friendly companion bot is no longer trapped in the past, finally closing the knowledge gap with competitors.
New York City just became the first major U.S. city to drop an Artificial Intelligence Action Plan, outlining 37 steps to weave AI into municipal services while guarding against algorithmic bias. As part of the rollout, the city launched a new chatbot on its MyCity Business portal to help entrepreneurs navigate red tape. It's an ambitious bid to cut bureaucracy without letting rogue algorithms run city hall.
Stack Overflow just laid off 28 percent of its staff—over 100 employees—primarily targeting its go-to-market and support teams. As developers increasingly turn to AI chatbots for code assistance, the veteran coding forum is feeling the heat just a year after doubling its headcount to chase profitability.
The SEC's Gary Gensler just warned that an AI-triggered financial crisis is "nearly unavoidable" by the late 2020s or early 2030s. The danger isn't necessarily one reckless trader, but entire Wall Street institutions quietly relying on the same underlying foundation model or data aggregator. Why it matters: Regulators are built to police individual banks, making systemic AI monocultures a massive blind spot.
Descript just launched Underlord, an agentic AI video and podcast co-editor designed to take tedious editing off your hands. It can analyze your script and video, center speakers, add captions, and clean up retakes from plain prompts. Why it matters: You can now direct an AI like a human collaborator instead of manually slicing timelines.
Portkey just revealed that GPT-4 latency has steadily dropped over the past three months, effectively cutting 99th-percentile wait times by more than half. While complex prompts still spike processing, median latencies now sit comfortably below 1ms per token. It is still pricier, but speed is no longer the bottleneck holding teams back.
Leaked screenshots reveal Google is quietly bringing its multimodal Gemini model to MakerSuite alongside an unannounced tool called "Stubbs." The secret feature lets creators generate, deploy, and share functional app prototypes with just a single text prompt and an optional reference image.
Google just promised to defend Cloud and Workspace customers against intellectual property lawsuits. The "two-pronged" indemnity covers both copyrighted training data and generated outputs across tools like Vertex AI and Duet AI. Why it matters: Enterprise users get peace of mind, though Google warned it won’t shield anyone intentionally generating infringing work.
Masterpiece Studio and NVIDIA just rolled out Masterpiece X – Generate, a browser-based text-to-3D playground powered by NVIDIA Picasso. Creators can turn simple text prompts into rigged and textured 3D models of humans, animals, or objects, then export them directly into Blender, Unity, or Unreal Engine. While it is not meant for AAA game assets, it dramatically lowers the barrier for rapid 3D prototyping.
ElevenLabs just teamed up with video generation platform Pictory AI to bring its realistic voice synthesis to automated video editing. Creators can now turn text scripts and blog posts into videos with human-sounding narration, initially focusing on English and German content.
Spawning just launched Kudurru, a defense network designed to block AI scrapers from harvesting artists’ works. Artists can automatically block scrapers crawling datasets like LAION-5B or actively poison their training data by returning junk images—like a middle finger. Why it matters: Until copyright laws catch up, creators finally have a free, active way to push back.
OpenAI just overhauled the core values on its careers page, putting "AGI focus" at the very top of the list. The new guidelines bluntly state that "anything that doesn’t help with that is out of scope," alongside traits like "Scale" and "Intense and scrappy." It marks a clear cultural shift as the company narrows its entire mission toward racing to human-level AI.
SAG-AFTRA just endorsed the "No Fakes Act," a new bipartisan discussion draft presented in the U.S. Senate. The proposal aims to establish federal protections prohibiting the use of AI-generated voice or likeness replicas without explicit informed consent. Why it matters: It directly addresses the union's core strike demands, blocking studios from cutting costs by replacing background actors and voice performers with synthetic AI replicas.
Google just added image generation and drafting tools directly into its Search Generative Experience (SGE). Users can create up to four AI images from a text prompt, tweak the descriptions, or draft messages to export straight into Gmail and Google Docs. It turns search from a link directory into an active workspace, though it's currently limited to US SGE testers.
OpenAI is reportedly planning major developer updates at its upcoming November 6 conference to slash costs by up to 20 times. The upgrades include a memory-enabled stateful API to curb token usage and vision capabilities for analyzing images. Why it matters: OpenAI wants to cement its platform as the default infrastructure before startups defect to open-source alternatives.
Character.AI just rolled out Character Group Chat, allowing c.ai+ subscribers to chat with multiple AI personas and human friends in the same room. Users can run text adventure games, host multi-character book clubs, or debate historical figures simultaneously. The feature is currently live on iOS and Android apps, with web support coming soon.
Researchers from Harvard and Oxford just introduced EVEscape, an AI tool designed to forecast how viruses will mutate before new variants actually emerge. By combining evolutionary sequences with structural biology, the system spots dangerous escape mutations in SARS-CoV-2 and flu, helping scientists design future-proof vaccines.
Mozilla just started testing Review Checker in Firefox, a built-in tool powered by Fakespot that uses AI to spot fake product reviews. It grades review authenticity from A to F across sites like Amazon, BestBuy, and Walmart, while using Oblivious HTTP to keep browsing data private. A full rollout is slated for November.
Adobe just showcased 11 experimental AI prototypes during its MAX Sneaks session, spanning video, audio, 3D, and design. Highlights include Project Fast Fill, which brings generative fill to video frames via text prompts, and Project Primrose, a dynamic fabric dress that transforms patterns in real time. While these are research projects, they offer a glimpse into the next generation of creative editing tools.
Replit just dropped Replit Code V1.5 3B on Hugging Face, an open-source completion model trained on 1 trillion tokens across 30 programming languages. Despite its compact 3B size, fine-tuning lets it outperform larger models like CodeLlama 7B. Best of all, it is free for commercial fine-tuning without strict restrictions.
Adobe and the C2PA coalition just launched a new "icon of transparency" to reveal whether media was created with AI. Hovering over the badge shows who made it and which tools were used, functioning like a nutrition label to counter deceptive deepfakes.
TikTok just launched Direct Post, letting creators publish videos directly from editing tools like Adobe Premiere Pro, Adobe Express, and CapCut. Editors can set captions and audience privacy without leaving their workspace. Why it matters: It cuts out the tedious export-and-upload step, smoothing out AI-assisted video workflows.
ElevenLabs just rolled out AI Dubbing, a voice translation tool that converts speech into over 20 languages in minutes while keeping the original speaker's exact voice, emotion, and tone. It also automatically distinguishes multiple speakers and filters out background noise.
Google just expanded its suite of climate AI tools targeting floods, wildfires, and urban heat. The tech giant brought its Flood Hub river forecasting to the U.S. and Canada, partnered with the U.S. Forest Service on its biggest fire spread model update in 50 years, and scaled its Tree Canopy tool to over 2,000 cities worldwide. It is a concrete push to turn satellite data into life-saving emergency alerts.
Warren Buffett's longtime partner Charlie Munger just threw cold water on the generative AI boom, calling the technology overhyped. Speaking at Zoomtopia, the 99-year-old investor pointed out that software creating software has been around for decades, bluntly adding that AI won't cure cancer or solve all our problems.
AMD just signed an agreement to acquire Nod.ai, an open-source AI software startup behind projects like SHARK and Torch-MLIR. The team specializes in automated compiler tech that speeds up model deployment across AMD GPUs and chips. Why it matters: software has long been AMD's weak link against Nvidia, and this deal shows it's doubling down on open-source tools to catch up.
CCS Insight just forecasted a major reality check for generative AI in 2024 as runaway computing costs and regulatory hurdles replace peak hype. Running these massive models is becoming far too expensive for anyone outside Big Tech. Why it matters: The boom is hitting reality, bringing slower enterprise rollout, search watermarks, and the first deepfake arrests.
Adobe just unveiled Firefly Image 2 at Adobe MAX, bringing photorealistic detail, camera-like controls, and style-matching capabilities. The company also debuted the first generative AI model for vector graphics inside Illustrator and rolled out text-to-template in Express. Commercial safety remains the core pitch.
Disney just found itself under fire from the creative community over a Loki Season 2 poster suspected of using generative AI in its background. While artists flagged artifacts in the Shutterstock stock image, Shutterstock later investigated and insisted the piece was crafted with traditional software rather than AI tools.
Google Cloud just introduced new generative AI search capabilities built specifically for healthcare workers to pull patient data across fragmented clinical notes and records in one place. The system directly links back to original internal sources to prevent hallucinations and cut down clerical paperwork. Why it matters: Administrative overload burns out over half of physicians; letting doctors ask plain-language questions across messy files could reclaim hours of clinical time.
Google is developing an "Organize Tabs" feature for Chrome that automatically reorders open tabs and places them into organized groups. Located near tab search in the top-left corner, it aims to automate the decluttering process for users with too many open tabs.
OpenAI is reportedly exploring developing its own AI chips to reduce its reliance on Nvidia. As GPU shortages worsen and inference costs surge, CEO Sam Altman has made securing chips a top priority. Options under evaluation include in-house design and acquiring a chipmaker, though custom silicon could take years and billions to materialize.
OpenAI CTO Mira Murati just revealed that the company was completely unprepared for ChatGPT's viral explosion, noting all preparations became irrelevant within a few hours of launch. While the boom kicked off a global AI arms race, Murati warned that rival tech giants scrambling to ship products risk a dangerous race to the bottom on safety.
OpenAI is exploring building its own custom AI chips to tackle the GPU crunch and reduce its reliance on Nvidia, according to Reuters. CEO Sam Altman has reportedly made acquiring more chips a top priority, with options including in-house design or acquisitions. Why it matters: Running models like ChatGPT costs tens of billions of dollars, but custom silicon is a risky bet that takes years.
Meta FAIR and Inria just open-sourced Stable Signature, a new watermarking method for open-source AI image generators. Instead of slapping watermarks on exported images, it roots an invisible signature directly inside the model's decoder, making it survive cropping, compression, and ControlNet fine-tuning. Why it matters: it virtually eliminates false positives while tracing deepfakes straight back to the model that made them.
Meta's new Emu-powered sticker generator is already dodging its own guardrails. While direct sensitive prompts get blocked, simple synonyms and creative phrasing easily churn out kids with weapons, cartoon nudity, and graphic political caricatures. It is a stark reminder of how porous AI content moderation still is when rolled out to billions of users.
Google just unveiled Assistant with Bard, blending its generative AI chatbot directly into its digital helper. It works across voice, text, and images, connects with Gmail and Docs, and features a mobile overlay that sees what is on your screen. It is rolling out to select testers first before launching publicly on Android and iOS.
SoftBank CEO Masayoshi Son just declared that artificial general intelligence will surpass the collective intelligence of all humanity tenfold by the end of the decade. Speaking at SoftBank World, Son predicted AGI will unlock full autonomous driving and Nobel-grade discoveries, blunt warning that those who ignore AI risk becoming intellectual goldfish compared to those who embrace it.
AI is coming for traditional insurance underwriting, tackling painful paper-pushing, pricing miscalculations, and human errors. Machine learning models can analyze unstructured data like text, speech, and images to sharpen risk assessments and flag fraud early. Instead of replacing human underwriters, automated workflows handle low-complexity grunt work so staff can tackle complex cases and retain clients.
Meta is rolling out generative AI features to all advertisers inside Ads Manager, letting brands generate custom image backgrounds and spin up alternate ad copy in seconds. Following a limited test in May, the tech giant is directly upgrading its core ad-revenue engine, with the full rollout expected to finish next year.
The Browser Company just launched Arc Max, a new bundle of AI features for its Arc browser powered by GPT-3.5 and Anthropic. Instead of shoehorning a clumsy chat sidebar onto every page, Arc is adding five subtle, time-saving tools like automatic tab and download renaming, 5-second link previews on hover, and page-specific Q&A via Command-F. All features are opt-in, with users voting on what stays after 30 days.
University of Maryland researchers just revealed that every existing AI image watermarking method can be cracked. Attackers can easily wipe invisible watermarks clean or even inject them into human-made photos to trigger false positives. While tech giants are betting big on watermarking to fight deepfakes, it's far from a silver bullet.
Google DeepMind just partnered with 33 academic labs to launch Open X-Embodiment, pooling data across 22 robot types, alongside new RT-X models. Cross-embodiment training boosted task success rates by 50% on average and tripled emerging skills on RT-2-X. It marks a major push to train generalist robots on shared data, and both the dataset and RT-1-X weights are now public.
Stability AI just released an experimental version of Stable LM 3B, a compact 3-billion-parameter language model tailored to run locally on laptops and handhelds. Despite its small footprint, it outperforms prior 3B models and even beats some 7B open-source models on common reasoning benchmarks. The base model is open-sourced under CC-By-SA 4.0 on Hugging Face, with an instruction-tuned variant coming soon.
Google just introduced Chromebook Plus, a new hardware category that bakes system-level AI tools into affordable laptops starting at $399. Built with guaranteed minimum specs, the devices offer OS-level noise reduction and lighting tweaks for any video calling app, alongside Google Photos Magic Eraser. Why it matters: Google is quietly turning budget computers into capable, AI-first machines without inflating prices.
Visa just launched a $100 million generative AI ventures initiative through Visa Ventures to back startups shaping the future of payments. While most generative AI attention has centered on tasks and content creation, Visa is betting the tech is about to fundamentally reshape how global commerce works.
Humane just showcased its upcoming Ai Pin on the runway at Paris Fashion Week, pinned onto clothing worn by Naomi Campbell and other Coperni models. The appearance offered the first full look at the rounded square hardware. However, no one actually operated the device on stage, leaving interaction details and standalone capabilities unanswered.
