Open Source - Page 17 of 21
Open-source AI projects, community innovations, collaborative development, and freely accessible AI tools and frameworks.
Open-source AI projects, community innovations, collaborative development, and freely accessible AI tools and frameworks.
Everyone wants a good Christmas card. Almost no one wants to design one. This is the annual dilemma that a few new AI tools are trying to solve. Their pitch is simple. Upload a selfie. Write a message. Choose some festive junk from a menu.
Global Capability Centres are desperately short on data scientists. StarRez, a student housing software company, thinks the answer isn't just to hire them. It's to build a whole new team from nothing in Hyderabad.
ESG data is messy. It lives in PDFs, APIs, scattered databases, and inconsistent formats.
The hype is over. After years of breathless promises and billions in investment, the architects of the AI boom are publicly dismantling their own flagship goal.
Xiaomi is selling an AI model for less than the price of a text message. For a tenth of a cent per million input tokens, you can rent MiMo-V2-Flash. It is fast and it is cheap, which is exactly the point. The numbers are a direct challenge.
OpenAI just wrote Amazon a check for ten billion dollars. The money buys them a future that does not depend entirely on NVIDIA. The deal, part of a larger $38 billion partnership, secures access to Amazon's custom Trainium3 chips.
Zepto Cafe's new system doesn't order your coffee through a secret business deal. It does something more straightforward and a little weirder. It makes a computer open a web browser and click 'add to cart' for you. The process splits the work.
The headline screams contradiction, but Swen Vincke is adamant: generative AI hasn’t made Larian more efficient. It hasn’t replaced a single writer or artist on *Divinity*.
The Hindi-speaking world has waited for an AI that doesn’t just translate, it thinks in the same tongue.
Nvidia is moving from selling the shovels to owning the mine. Its new Nemotron 3 models have scored near the top of public benchmarks for downloadable, modifiable AI. The performance is strong. The strategy is clearer.
Forget raw compute. Researchers behind the new LongCat-Image model have a different credo: cleaner data beats bigger models.
HBO Max just launched twelve permanent channels. You can now watch Friends or Game of Thrones exactly as a 1998 college dorm did: aimlessly, and with no remote in sight. It’s a strange, old-fashioned solution to modern choice paralysis.
A forty-dollar paperweight for a cable is an insult. It's also the only product that gets why you'd spend the money. The Weighter is a small, heavy block. It grips your expensive table without adhesive and holds charging cables in place.
The wait is over. Google’s Gemini AI is no longer an Android-exclusive trick inside Chrome. Starting today, iPhone and iPad users can summon a smart assistant directly from the address bar, no app switching, no extra steps.
A machine crushed the Putnam exam, and then another did nearly the same. The era of humans dominating elite math is finished. DeepSeekMath-V2 posted a 118 out of 120, annihilating the historical human record.
The terminal has always been the coder’s quiet sanctuary, until now. Mistral AI just turned that sanctuary into a command center with the release of DevStral 2 (123B parameters), DevStral Small 2 (24B), and the Mistral Vibe CLI.
$40 billion. That's the price tag corporations have slapped on generative AI in a single year. The staggering sum is meant to buy transformation, efficiency, and a leg up on the competition. Instead, for 95% of companies, it's buying nothing.
India's government is done watching Google and OpenAI feast. A new committee report, delivered to the Department of Science and Technology, demands a license for every byte of Indian data used to train AI. It insists on payment for every scrap.
Intel is snapping up SambaNova Systems in what looks, for all the world, like a clearance sale. Remember that heady $5 billion valuation from 2021? Gone. Reports say the price will be lower. The connections here are intensely personal.
The 24-billion-parameter model is back, and this time, it’s built to run on your laptop. Mistral’s Devstral 2 arrives as a direct answer to a persistent question in AI: how do you deliver real coding autonomy without a data center?
Learn to build AI-powered apps without coding. Our comprehensive review of No Code MBA's course.
Curated collection of AI tools, courses, and frameworks to accelerate your AI journey.
Get the week's most important AI news delivered to your inbox every week.