Latest AI News
How India’s 50th Largest City Plans to Become a Global Quantum Hub
IIIT Dharwad’s CoE aims to bring quantum access, training, and industry links to tier-2 regions within months.
View
Oscars Revise Eligibility Rules, Ban AI Actors and Scripts From Winning Awards
The Academy Awards, better known as the Oscars, are tightening the rules for artificial intelligence (AI). New rules, released on Friday, specifically highlight that AI actors or screenplays written using an AI tool will not be eligible to be nominated or win the award starting next year. Additionally, the organisation behind the Oscars can also request more information from filmmakers about the usage of the technology and human authorship. The development comes as AI actors such as Tilly Norwood have gained popularity in the mainstream.
View
How Neoclouds Are Finally Unlocking AMD’s AI GPUs
AMD is leaning on highly specialised neocloud startups like MangoBoost and TensorWave to make its chips enterprise-ready.
View
Applied Materials to Expand Chip Packaging with NEXX Acquisition
The deal aims to strengthen panel-level packaging capabilities for AI chip production and accelerate the adoption of large-format technologies.
View
Pixxel, Sarvam Partner to Launch Orbital Data Centre Satellite
The Pathfinder satellite, in the 200 kg class, will carry data centre-grade GPUs typically used in terrestrial AI systems.
View
15 Indian Govt Leaders Leading AI-Backed Public Sector Transformation
From Cabinet ministers to mission CEOs — the officials shaping IndiaAI, defence tech, digital health, and the nation's ₹10,372 crore artificial intelligence ambition.
View
In Harvard study, AI offered more accurate emergency room diagnoses than two human doctors
A new study examines how large language models perform in a variety of medical contexts, including real emergency room cases — where at least one model seemed to be more accurate than human doctors. The study waspublished this week in Scienceand comes from a research team led by physicians and computer scientists at Harvard Medical School and Beth Israel Deaconess Medical Center. The researchers said they conducted a variety of experiments to measure how OpenAI’s models compared to human physicians. In one experiment, researchers focused on 76 patients who came into the Beth Israel emergency room, comparing the diagnoses offered by two internal medicine attending physicians to those generated by OpenAI’s o1 and 4o models. These diagnoses were assessed by two other attending physicians, who did not know which ones came from humans and which came from AI. “At each diagnostic touchpoint, o1 either performed nominally better than or on par with the two attending physicians and 4o,” the study said, adding that the differences “were especially pronounced at the first diagnostic touchpoint (initial ER triage), where there is the least information available about the patient and the most urgency to make the correct decision.” In Harvard Medical School’spress releaseabout the study, the researchers emphasized that they did not “pre-process the data at all” — the AI models were presented with the same information that was available in the electronic medical records at the time of each diagnosis. With that information, the o1 model managed to offer “the exact or very close diagnosis” in 67% of triage cases, compared to one physician who had the exact or close diagnosis 55% of the time, and to the other who hit the mark 50% of the time. “We tested the AI model against virtually every benchmark, and it eclipsed both prior models and our physician baselines,” said Arjun Manrai, who heads an AI lab at Harvard Medical School and is one of the study’s lead authors, in the press release. To be clear, the study didn’t claim that AI is ready to make real life-or-death decisions in the emergency room. Instead, it said the findings show an “urgent need for prospective trials to evaluate these technologies in real-world patient care settings.” The researchers also noted that they only studied how models performed when provided with text-based information, and that “existing studies suggest that current foundation models are more limited in reasoning over nontext inputs.” Adam Rodman, a Beth Israel doctor who’s also one of the study’s lead authors,warned the Guardianthat there’s “no formal framework right now for accountability” around AI diagnoses, and that patients still “want humans to guide them through life or death decisions [and] to guide them through challenging treatment decisions.” Ina post about the study, Kristen Panthagani, an emergency physician, said this is an “an interesting AI study that has led to some very overhyped headlines,” especially since it was comparing AI diagnoses to those from internal medicine physicians, not ER physicians. “If we’re going to compare AI tools to physicians’ clinical ability, we should start by comparing to physicians who actually practice that specialty,” Panthagani said. “I would not be surprised if a LLM could beat a dermatologist at an neurosurgery board exam, [but] that’s not a particularly helpful thing to know.” She also argued, “As an ER doctor seeing a patient for a first time, my primary goal isnotto guess your ultimate diagnosis. My primary goal is to determine if you have a condition that could kill you.” This post and headline have been updated to reflect the fact that the diagnoses in the study came from internal medicine attending physicians, and to include commentary from Kristen Panthagani.
View
‘This is fine’ creator says AI startup stole his art
You’ve seen this comic before: An anthropomorphic dog sits smiling, surrounded by flames, and says, “This is fine.” It’s become one ofthe most durable memesof the past decade, and now AI startupArtisanseems to have incorporated it into an ad campaign — an ad for whichKC Green, the artist who created the comic, said his art was stolen. A Bluesky postseems to show an ad in a subway station featuring Green’s art, except the dog says, “[M]y pipeline is on fire,” and an overlaid message urges passersby to “Hire Ava the AI BDR.” Quoting that post,Green saidhe’s “been getting more folks telling me about this” and that “it’s not anything [I] agreed to.” Instead, he said the ad has “been stolen like AI steals,” and he told followers to “please vandalize it if and when you see it.” i've been getting more folks telling me about this and it's not anything i agreed to. it's been stolen like AI steals. please vandalize it if and when you see it. When TechCrunch sent Artisan an email asking about the ad, the company said, “We have a lot of respect for KC Green and his work, and we’re reaching out to him directly.” In a follow-up email, the company said it had scheduled time to speak with him. Artisan has courted controversy with its ads before, specifically withbillboards urging businesses to “Stop hiring humans”— although founder and CEO Jaspar Carmichael-Jack insisted that the message was about “a category of work,” not “humans at large.” “This is fine” first appeared in Green’s webcomic“Gunshow”in 2013, and while he hasn’t disavowed the smiling-melting dog entirely (he recentlyturned the comic into a game), it’s clearlyescaped from his control. And of course, Green is far from the only artist to see his meme-able art used in ways he finds objectionable. But some artists have still taken action when their art is monetized or used in commercial ways without their permission, for example when cartoonist Matt Furie sued right-wing conspiracy theory site Infowars for using his character Pepe the Frog in a poster. (Furie and Infowars eventually settled.) Green told TechCrunch via email that he will be “looking into [legal] representation, as I feel I have to.” Still, he said it “takes the wind out of my sails” that he has to take “time out of my life to try my hand at the American court system instead of putting that back into what I am passionate about, which is drawing comics and stories.” Green added, “These no-thought A.I. losers aren’t untouchable and memes just don’t come out of thin air.”
View
Bengaluru-based GalaxEye Launches Mission Drishti, ‘World’s First’ OptoSAR Satellite
The satellite weighs 190 kg and is the largest privately developed Earth observation satellite from India.
View
When Kids Stop Playing Games and Start Coding Them
Children are moving from playing games to creating immersive experiences, thanks to the ease of Luau scripting offered by platforms like Roblox.
View
AI-generated actors and scripts are now ineligible for Oscars
The organization behind the Academy Awardsreleased new Oscar ruleson Friday, including several that address the use of generative artificial intelligence. The Academy of Motion Picture Arts and Sciences said that only performances “credited in the film’s legal billing and demonstrably performed by humans with their consent” will be eligible for Academy Awards. Similarly, the academy said that screenplays must be “human-authored” to be eligible. The academy also said it has the right to request more information about a film’s AI usage and “human authorship.” These rule changes come as an independent film is in the works withan AI-generated version of Val Kilmer, as AI “actress” Tilly Norwoodkeeps making headlines, and asnew video modelsare causing at least a few filmmakers to make sweeping declarations of despair. AI was also one of the main sticking points inthe actors’ and writers’ strikesback in 2023. Outside Hollywood, at least one novel has beenpulled by its publisherdue to the apparent use of AI, and other writers’ groups are declaring that AI usage makes workineligible for awards.
View
The best AI dictation apps, tested and ranked
AI dictation apps have come a long way in a short time. For years they were slow and inaccurate — unless you spoke with a particular accent and enunciated clearly. Advances in large language models (LLMs) and speech-to-text models have changed that, producing systems that can decipher speech more accurately while retaining enough context to format the text correctly. Developers have also built in features to automatically remove filler words, fix stumbles, and handle punctuation — outputting text that needs far fewer edits. With dozens of such apps now on the market, we’ve rounded up our picks for the best and most useful dictation apps available right now. Wispr Flowis a well-funded AI dictation app that lets you add custom words and instructions for dictation. It has native apps for macOS, Windows, and iOS; an Android version is in the works. The app lets you customize how it transcribes your text by choosing from “formal,” “casual,” and “very casual” styles for different kinds of writing, such as personal messaging, work, and email. And if you use it with vibe-coding tools like Cursor, you can turn on a feature to automatically recognize variables or tag files in the chat. The app lets you transcribe up to 2,000 words per week for free on desktop, and 1,000 words per month on iOS. Paid subscription plans offer unlimited transcription and start at $15 per month. Willowadvertises itself as a big time-saver for those who don’t like to type. Alongside common features like automatic editing and formatting, the app uses large language models to generate a full passage of text from just a few dictated words. Willow also takes a more privacy-focused approach by storing all transcripts locally on your device and lets you opt out of model training entirely. It also lets you add custom vocabulary to help it adapt to your industry’s terminology, or your local dialect. Willow lets you dictate 2,000 words per month on its desktop app for free. Individual subscription plans start at $15 per month, unlocking unlimited dictation and enabling the app to remember your writing style. If privacy if your priority,Monologuelets you download its AI model directly to your device for transcriptions, keeping your data off the cloud entirely. What’s more, the app lets you customize its tone depending on the app you use it with. Monologue lets you transcribe 1,000 words per month for free; a subscription costs $10 per month or $100 per year. The company also sends its most active users a physical shortcut device called the Monokey to use with the app. Keyboards are so 1983.You only need one key—Monokey, the limited edition device that turns your voice into text in Monologue.We're giving away 10, along with a free annual subscription to Monologue.pic.twitter.com/nXuz1ll2LU Superwhisperis primarily a dictation app, but it can also transcribe from audio or video files. The app lets you choose and download AI models, including several of its own at different speeds and accuracy levels, along with Nvidia’s Parakeet speech-recognition models. The app also lets you write custom prompts to steer the output, and you can view both processed and unprocessed transcripts directly from your system keyboard. The basic voice-to-text feature is free to use, and you get 15 minutes to test Pro features such as translation and transcription. The paid tier lets you use your own AI API keys and connect cloud and local models without any usage caps. The monthly plan costs $8.49 per month, the annual plan costs $84.99 per month, or you can pay $249.99 for a lifetime subscription. TheVoiceTyprapp takes an offline-first, no-subscription approach, letting you use local models for transcription. It also has aGitHub repositoryfor those who want to host and run the open source version themselves. VoiceTypr supports over 99 languages and works on both Mac and Windows. The app is available to try for three days for free, and after that, it will allow you to buy a lifetime license. The app costs $35 for one device, $56 for two, and $98 for four devices. Aquais a Y Combinator-backed voice-typing app for Windows and macOS that claims to be one of the fastest tools in the category in terms of latency (the delay between when you speak and when text appears on screen). Besides handling grammar and punctuation, Aqua also lets you autofill text by saying phrases — you can say “my address” and have Aqua type it in, for example. The app also offers its own speech-to-text API, letting other apps plug into Aqua’s transcription engine. The free tier gets you 1,000 words per month. Paid plans start at $8 per month bill annually and unlock unlimited words and 800 custom dictionary values. Handyis an open-source, free transcription tool that runs on Mac, Windows, and Linux. The app is pretty basic and doesn’t offer much customization, but if you want to start using your voice more and don’t want to pay, it is a good option. The app has a basic settings menu that lets you toggle push-to-talk and change the hotkey to activate transcription. Typelessstands out for its high free word count. The company claims it doesn’t retain any data or use it to train AI models. Typeless also offers to rewrite sentences you may have fumbled. The app lets you dictate up to 4,000 words per week (roughly 16,000 words per month) on its free tier. You can pay $12 per month (billed annually) to unlock unlimited words and get access to new features. Typeless is available for Windows and macOS only. VoiceInkis anopen-sourceprivate dictation app for Mac. The app supports global shortcuts for recording start/stop, along with a push-to-talk mode. It reads the context on screen and adjusts its output accordingly. The app can automatically detect certain apps and URLs and apply custom formatting or rules to each. It also has an assistant mode that can answer your questions. The app costs $25 for lifetime access for one device, $39 for two devices, and $49 for three devices. Dictatois a dictionary app for Mac priced at €9.99 — roughly $12 — that gives you lifetime access and two years of feature updates. The app works with offline models like Parakeet, Whisper, and Apple Speech Analyzer, and uses Apple Intelligence for light reading and filler word removal. Thanks to these local models, the app claims a super fast 80ms latency, meaning text appears almost instantly after you speak. AudioPenbegan asa web-based voice notes app, but it has evolved over the years. Its Mac version now lets you dictate text and rewrite it in your preferred format and style, switching between different styles at any time. Besides live transcription, AudioPen allows you to store audio notes across platforms, combine notes for summaries, upload audio files, and rewrite existing notes using AI. The app costs $33 for three months, $99 for a year, and $159 for two years.
View
