Latest Tech News

Stay updated with the latest in technology, AI, cybersecurity, and more

Filtered by: mod Clear Filter

Evaluating LLMs for my personal use case

Most models are excellent, so cost and latency dominate. It’s great that AI can win maths Olympiads, but that’s not what I’m doing. I mostly ask basic Rust, Python, Linux and life questions. So I did my own evaluation. I gathered 130 real prompts from my bash history (I use command line tool llm). I had Qwen3 235B Thinking and Gemini 2.5 Pro group them into categories. They both chose very similar ones, broadly (with examples): Programming - “Write a bash script to ..” Sysadmin - “With curl

ChatGPT-5 Impressions: Fast, but a Bit Impersonal

ChatGPT-5 likes getting straight to the point. For some, that's a reprieve from its chattier predecessor, GPT-4o. For others, something will certainly seem off. Despite the hype leading up to OpenAI's launch of GPT-5, it ultimately doesn't feel too different from 4o. The quality of responses in my ongoing testing seems to be at the level of past models, including the o3 "reasoning" model. The major difference is that some responses generate very quickly in relatively few words, while others ca

OpenCUA’s open source computer-use agents rival proprietary models from OpenAI and Anthropic

Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now A new framework from researchers at The University of Hong Kong (HKU) and collaborating institutions provides an open source foundation for creating robust AI agents that can operate computers. The framework, called OpenCUA, includes the tools, data, and recipes for scaling the development of computer-use agents (CUAs). Models trained usin

Apple trained a large language model to efficiently understand long-form video

Apple researchers have developed an adapted version of the SlowFast-LLaVA model that beats larger models at long-form video analysis and understanding. Here’s what that means. The nerdy bits Very basically, when an LLM is trained to also understand video, it learns to split videos into frames, apply computer vision to extract visual features, analyze how those features change over time, and align all of that with language so it can describe or reason about the video in the form of text. One v

College student’s “time travel” AI experiment accidentally outputs real 1834 history

A hobbyist developer building AI language models that speak Victorian-era English "just for fun" got an unexpected history lesson this week when his latest creation mentioned real protests from 1834 London—events the developer didn't know had actually happened until he Googled them. "I was interested to see if a protest had actually occurred in 1834 London and it really did happen," wrote Reddit user Hayk Grigorian, who is a computer science student at Muhlenberg College in Pennsylvania. For t

MCP-Universe benchmark shows GPT-5 fails more than half of real-world orchestration tasks

Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now The adoption of interoperability standards, such as the Model Context Protocol (MCP), can provide enterprises with insights into how agents and models function outside their walled confines. However, many benchmarks fail to capture real-life interactions with MCP. Salesforce AI Research developed a new open-source benchmark it calls MCP-Un

Apple reportedly in talks with Google to make Gemini the foundation of Siri

Ryan Haines / Android Authority TL;DR Apple reportedly approached Google to explore developing a custom AI model to serve as Siri’s foundation. This AI model may already be in training to run on Apple’s servers. The Cupertino-based firm is said to be developing a version of Siri powered by its models and a second version that runs on third-party technology. Apple has long been trailing behind its competitors in the AI space. In fact, the company was expected to launch a revamped version of S

Apple considers Google Gemini to power next-gen Siri, internal AI ‘bake-off’ underway

Apple seems open to anything and everything when it comes to delivering the next generation of Siri. After reports that it could be powered by OpenAI or Anthropic, Google has entered the conversation. Mark Gurman reports for Bloomberg that Google, which offers a ChatGPT competitor called Gemini, is actually training a model that could run on Apple’s servers to power the new Siri experience: The iPhone maker recently approached Alphabet Inc.’s Google to explore building a custom AI model that w

Deals: Amazon Labor Day sale, most affordable 24GB M4 MacBook Pro $300 off, Apple Watch Ultra 2 $150 off, more

Today’s 9to5Toys Lunch Break deals are kicking off with the now live Amazon Labor Day sale where you’ll find plenty of big-time Apple gear deals. All of the details you need on that are right here, but we are also once again tracking the 24GB/512GB M3 MacBook Air for $400 less than the M4 model (it is now back in stock) alongside the most affordable M4 MacBook Pro with 24GB of RAM at $300 off, and Apple Watch Ultra 2 models at $150 off the list prices. All of that and more awaits below. Looking

GPT-5 usage limitations: what are they, how does this compare to GPT-4 family?

Edgar Cervantes / Android Authority GPT-5 arrived a few weeks ago, though its rollout hasn’t been entirely smooth. While the model shows plenty of promise, its debut also meant the abrupt removal of every other GPT model from ChatGPT’s user-facing UI. Since then, some old models have returned, and there have been a few other changes to the way the system works. Furthermore, many of the initial GPT-5 usage limits have been temporarily enhanced since launch. Let’s dive in and take a closer look

Tesla Will Use a Powerful New Weapon in AI Race

Tesla will partner up with AI darlings DeepSeek and Bytedance’s Doubao on new tools in its cars in China, according to a document uploaded to Tesla’s official website. Bloomberg reports that Doubao will work on voice command-related tools like the temperature in a Tesla, navigation, and in-car entertainment, while DeepSeek will handle the AI side of things. The move may be a way for Tesla to boost its Chinese deliveries, which dropped 8.4% from December to June compared to the same time period

24,000-Watt Scooter Is Going for a 100 MPH Speed Record at Bonneville

Get The Drive’s daily newsletter The latest car news, reviews, and features. Email address Sign Up Thank you! Terms of Service & Privacy Policy. A British outfit called Bo makes a sleek little electric scooter called the Model-M. The best variant claims a 22 mph top speed and a 40-mile range for $2,500. Sounds pricey but practical. The prototype Turbo model is another story—it’s gunning for triple-digit speeds at the Bonneville salt flats next month. Bo has been around since 2022, but seems to

Don’t sleep on Cohere: Command A Reasoning, its first reasoning model, is built for enterprise customer service and more

Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now I was in more meetings than usual today so I just caught up to the fact that Cohere, the Canadian startup geared co-founded by former Transformer paper author Aidan Gomez toward making generative AI products work easily, powerfully, and securely for enterprises, has released its first reasoning large language model (LLM), Command A Reasonin

Google's AI Mode Is Getting New Agentic Features for Restaurant Reservations and More

Google's AI Mode is getting new agentic capabilities and expanding to 180 new countries and territories around the world, Google announced Thursday. AI Mode allows you to ask questions directly to Google and it'll kick off a series of searches on your behalf. From there, it will surface relevant information to your query without the need to do any of the deep research yourself. It changes the way you search for things online, and it's getting even smarter with this latest rollout. Powered by De

Building AI products in the probabilistic era

I was recently trying to convince a friend of mine that ChatGPT hasn't memorized every possible medical record, and that when she was passing her blood work results the model was doing pattern matching in ways that even OpenAI couldn't really foresee. She couldn't believe me, and I totally understand why. It's hard to accept that we invented a technology that we don't fully comprehend, and that exhibits behaviors that we didn't explicitly expect. Dismissal is a common reaction when witnessing A

Chan Zuckerberg Initiative’s rBio uses virtual cells to train AI, bypassing lab work

Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now The Chan Zuckerberg Initiative announced Thursday the launch of rBio, the first artificial intelligence model trained to reason about cellular biology using virtual simulations rather than requiring expensive laboratory experiments — a breakthrough that could dramatically accelerate biomedical research and drug discovery. The reasoning mod

Google's AI Mode can now find restaurant reservations for you - how it works

SvetaZi/iStock/Getty Images Plus via Getty Images ZDNET's key takeaways Google's AI Mode can now suggest restaurant reservations. Users can also now receive personalized dining recommendations. Google is aiming to build a more dynamic and personalized web. AI Mode can now find restaurant reservations, Google announced in a blog post Thursday. Users can now request reservations in local restaurants according to constraints like party size and dietary restrictions, and the system will automat

Building AI Products in the Probabilistic Era

I was recently trying to convince a friend of mine that ChatGPT hasn't memorized every possible medical record, and that when she was passing her blood work results the model was doing pattern matching in ways that even OpenAI couldn't really foresee. She couldn't believe me, and I totally understand why. It's hard to accept that we invented a technology that we don't fully comprehend, and that exhibits behaviors that we didn't explicitly expect. Dismissal is a common reaction when witnessing A

I Gave GPT-5 Some Work Tasks in Copilot. Here's What's New

OpenAI's latest GPT-5 model promises to be the most advanced AI assistant yet, but for the millions of workers using Microsoft Copilot for their daily routines, the reality is a bit more nuanced. Microsoft announced last week that GPT-5, OpenAI's latest ChatGPT AI model, is now integrated across its Copilot services, including Microsoft 365 Copilot, Copilot Studio, GitHub Copilot and the standalone Copilot app. While GPT-5 does deliver some noticeable improvements in polish, detail and helpfuln

Meet Surya, a New AI Model From NASA and IBM That Can Predict Solar Flares

Today's solar flare forecasting is pretty straightforward. Earth has several instruments pointed at the sun that monitor it, and when a solar flare erupts, NOAA predicts whether it'll hit Earth and reports it via the Space Weather Prediction Center. But NASA and IBM may be able to do this faster and with a little more accuracy thanks to a new artificial intelligence model named Surya. Surya, which is Sanskrit for "Sun," is a 366M-parameter AI model developed to analyze the various cool things t

Honor's Magic V5 Boasts On-Device Live AI Call Translation for Guaranteed Privacy

"Hola! ¿Hablas inglés?" I asked the woman who answered the phone in the Barcelona restaurant. I was calling in a futile attempt to make a reservation for the CNET team dinner during Mobile World Congress this year. Unfortunately, I don't know Spanish (I learned French and German at school). And as it turned out, she didn't speak English either. "No!" she said, and brusquely hung up. What I needed in that moment was the kind of AI call translation feature that's becoming increasingly prevalent

Microsoft AI chief says it’s ‘dangerous’ to study AI consciousness

AI models can respond to text, audio, and video in ways that sometimes fool people into thinking a human is behind the keyboard, but that doesn’t exactly make them conscious. It’s not like ChatGPT experiences sadness doing my tax return… right? Well, a growing number of AI researchers at labs like Anthropic are asking when — if ever — might AI models develop subjective experiences similar to living beings, and if they do, what rights should they have? The debate over whether AI models could on

ChatGPT-5 Lets You Choose Your AI Model. These Are Your Options

The biggest pushback after OpenAI announced its new GPT-5 model for ChatGPT came from devotees of older models who felt the new generative AI chatbot lacked the panache of its predecessors. Now you have more choices of pre-GPT-5 models (although you'll have to hunt for some of them) and better control over which components of GPT-5 handle your questions. OpenAI is still sorting through a somewhat rocky launch of GPT-5, led by complaints about the lack of model choices. The model has been antic

Google AI Mode is expanding to 180 countries and adding an agentic restaurant finder

Google’s seemingly unrelenting quest to infuse AI into every aspect of your online life just got a lot more global in scope, with the company expanding its AI Mode in Search to over 180 new countries . AI Mode has previously only been available in the US, India and the UK, and while English remains the only supported language right now, Google says it’ll add more soon. Google is also expanding its AI Mode’s agentic capabilities, so you can now use natural language to find restaurant reservation

Google Search can now present real-time table booking results, but only if you pay up for AI

Tushar Mehta / Android Authority TL;DR Google’s AI Mode is getting an agentic mode to help you with hotel reservations. AI Mode will also scan for your diet and seating preferences and offer suggestions for cafes and restaurants in particular areas. It is now expanding to over 180 countries for searches in English. Yesterday’s Pixel launch displays Google’s determination to bring agentic AI features to Android devices. While Magic Cue on Pixel 10 is just a step in that direction, the company

Google Search’s AI Mode is going global and getting smarter

is a news writer who covers the streaming wars, consumer tech, crypto, social media, and much more. Previously, she was a writer and editor at MUO. Posts from this author will be added to your daily email digest and your homepage feed. Google is putting AI Mode into Search in 180 countries and territories, allowing the AI-powered search feature to expand beyond the US, UK, and India. The expansion means more people can search for topics using a chatbot-like interface, though it’s still only av

Google’s AI Mode expands globally, adds new agentic features

Google is launching a global expansion of AI Mode, its experimental feature that allows users to ask complex questions and follow-ups to dig deeper on a topic directly within Search, the company announced on Thursday. The tech giant is also bringing new agentic and personalized capabilities to the feature. As part of the expansion, Google is bringing AI Mode to 180 new countries in English. Up until now, it’s only been available to users in the U.S., U.K., and India. Google plans to bring the f

Topics: ai google mode tech users

The Pixel 10 series brings even more panorama mode upgrades

Adamya Sharma / Android Authority TL;DR The Pixel 10 phones bring several more upgrades to the panorama mode. You can now take panoramas with the 5x camera, although ultrawide panoramas are still out of the question. Google has also brought manual controls to the panorama mode so you can fine-tune your snap. Google neglected panorama mode on its Pixel phones for a long time. However, the Pixel 9 series delivered a variety of welcome upgrades last year. Now, Google has brought even more addit

AI Mode in Search gets new agentic features and expands globally

AI is making Google Search radically more helpful, so you can ask any question on your mind and get things done. Starting today, we’re bringing more advanced agentic and personalized capabilities to AI Mode so you can make progress on your tasks and get more tailored information based on your interests. We’re also bringing AI Mode to even more people around the world. Read on for more. Get things done with agentic capabilities in AI Mode New agentic capabilities in AI Mode can help you get thi

Coris (YC S22) Is Hiring

AI Engineer Location: SF Bay Area ( 4+ days in office ) Experience Level: 3–5+ years Stack: Python, PyTorch, ML, LLMs, Django Type: Full-time 🧠 About Coris Coris is building the AI-first trust layer for global commerce. We partner with leading platforms, marketplaces, payment providers, and banks to transform how small business onboarding, monitoring, and lifecycle decisions are made - using AI on the ground to drive faster, smarter actions with less friction. One of our customers describ