Latest Tech News

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

Filtered by: questions Clear Filter

Why We Spiral

Say you’re a senior member of your team at work. You’re 12 minutes late to the weekly staff Zoom. Once you’ve “joined audio,” the first thing you hear is your old friend’s voice. “There you are! So glad you could fit us in.” You laugh and explain the disastrous traffic, difficult drop-off at your kids’ school, or whatever it was that messed up your morning. The moment passes and the conversation moves on. You turn to the job at hand, focused and ready to go. But what if you’re a junior staffer,

Close the loop: analytics that teach your chatbot to fix itself

Many chatbots stall for the same reason. Unanswered questions build up and nothing changes. Teams ship a release and move on. Users try again and give up. The way out is simple. Treat every miss as a signal. Capture it in a standard way. Decide whether it was noise or a real gap. Turn real gaps into small updates in guardrails or knowledge. Run that loop every week. Measure how fast it moves. Results improve without bigger models. Start with lean instrumentation Analytics only works if the tra

DeepCodeBench: Real-World Codebase Understanding by Q&A Benchmarking

At Qodo, we’ve created a new benchmark dataset of real-world questions derived from large, complex code repositories. We are excited to release the dataset, methodology, and prompts used in its creation to support further research and development. Motivation Enterprises often maintain massive codebases that are difficult for any individual developer to navigate and fully understand. Whether onboarding, doing routine development, or using AI-assisted workflows, teams often have questions about

‘Wicked: For Good’ Director Jon M. Chu Promises New Songs Will Hit All the Right Notes

As anticipation builds for Wicked: For Good, the second act to 2024’s Oscar-winning Wicked, director Jon M. Chu has revealed in a new interview with Entertainment Weekly that the upcoming film will feature not one, but two new bespoke songs—that’s one tune for each witch. While Chu didn’t announce the songs’ titles, the outlet teased that the new tracks by composer Stephen Schwartz are no “mere bids for a Best Original Song nomination.” In fact, Chu says the new songs are deeply needed addition

Google wants Gemini to keep the conversation going, and here’s how it’s going to do it

Mishaal Rahman / Android Authority TL;DR Google is testing a feature within Gemini that suggests follow-up questions. The suggestion prompts help users explore topics and engage in in-depth conversations to get more out of their AI interactions. The feature seems to be a limited test for now. If you’re old enough to have used search engines before their AI-fication, you’re most likely using your AI digital assistant more like a search engine and less like a digital assistant. That means more

Copilot's new File Explorer tricks are serious OneDrive time-savers - how to try them

Screenshot by Lance Whitney/ZDNET Follow ZDNET: Add us as a preferred source on Google. ZDNET's key takeaways You can access four new Copilot skills directly from File Explorer. You can summarize, ask questions, and compare up to five files. The process supports Microsoft 365 files, PDFs, and web files. Microsoft 365 users have long been able to turn to Copilot on the web to analyze and answer questions about their documents and other files. But now they can get help from Microsoft's AI di

This Amazon Lens upgrade lets you scan a product IRL and find it online in one click

Elyse Betters Picaro / ZDNET Follow ZDNET: Add us as a preferred source on Google. ZDNET's key takeaways Lens Live scans products for real-time shopping suggestions. Amazon's AI assistant can answer questions, offer details. The feature is currently available to some US users on iOS. Have you ever seen an item in a brick-and-mortar store and figured you could get it for a better price on Amazon, but couldn't string together the right keywords to find it online? Amazon Lens was designed to

An LLM is a lossy encyclopedia

Since I love collecting questionable analogies for LLMs, here's a new one I just came up with: an LLM is a lossy encyclopedia. They have a huge array of facts compressed into them but that compression is lossy (see also Ted Chiang). The key thing is to develop an intuition for questions it can usefully answer vs questions that are at a level of detail where the lossiness matters. This thought sparked by a comment on Hacker News asking why an LLM couldn't "Create a boilerplate Zephyr project sk

Kapa.ai (YC S23) is hiring research and software engineers

Why you should join kapa.ai We make it easy for technical companies to build AI assistants. Companies like Docker, Grafana and Mixpanel deploy kapa in the following ways: As chat interface on their public documentation to answer developer questions. As first line of defense on their support forms to reduce tickets. As internal assistant for their GTM teams to navigate their own complex product. We leverage companies existing technical knowledge sources including documentation, tutorials, fo

Show HN: SecretMemoryLocker – File Encryption Without Static Passwords

💾 SecretMemoryLocker (SecretML v2.23) Your personal digital vault – protected by your memories. 💡 Upcoming Feature: SecretML-Seed (SML-Seed) — your personal recovery key, coming soon and fully functional! 🚀 What's New in v2.23 — MirageLoop (SML-ML) Secret Memory Locker v2.23 introduces the unique MirageLoop (SML-ML) feature. This is not just an update — it’s a new reality of protection. 🔐 How it works When a wrong answer to a security question is entered — MirageLoop activates. to a sec

Join me and the Android Authority team for a Pixel 10 series AMA, August 27 at 1PM EST

Rita El Khoury / Android Authority After its announcement last week, the Pixel 10 series will officially be available for purchase on Thursday, August 28, 2025. Many of you have already pre-ordered one of the three phones, but many are probably still waiting for real-world feedback before they make up their mind. That’s what we’re here for. If you’re eager to know more about the Pixel 10, Pixel 10 Pro, and Pixel 10 Pro XL, their performance, how specific features work and how good they are, yo

Something Extremely Scary Happens When Advanced AI Tries to Give Medical Advice to Real World Patients

Image by Getty / Futurism Developments Last week, Google AI pioneer Jad Tarifi sparked controversy when he told Business Insider that it no longer makes sense to get a medical degree — since, in his telling, artificial intelligence will render such an education obsolete by the time you're a practicing doctor. Companies have long touted the tech as a way to free up the time of overworked doctors and even aid them in specialized skills, including scanning medical imagery for tumors. Hospitals ha

A summary of recent AI research (2016)

Story comprehension The robots of Westworld are not programmed solely by software developers. The bulk of the work is done by professional writers, who give each character a unique backstory. These stories give them the memories and depth they need to seem real to the park guests. When asked who they are, what they’ve done or why they feel a certain way, they can consult their backstory to find out the answer. Being able to answer questions about stories is a fundamental requirement for being

All Souls exam questions and the limits of machine reasoning

Oxford University is immersed in the past like no other place I’ve seen. One example: when I was a visiting student at Oxford in 2005, I remember meeting two students at a pub one evening. They were drinking ivy-laced beer. The reason, I was told, is that centuries ago, a student from Lincoln College had murdered a student of Brasenose. Ever since then, Brasenose students had been allowed into Lincoln and given free beer once a year. Here’s the event back in 1938: The actual truth behind “ivy

Doximity buys Pathway Medical for $63 million to help doctors get AI-powered answers

Doximity at the New York Stock Exchange for its initial public offering on June 24, 2021. Doximity is diving deeper into artificial intelligence, announcing on Thursday the acquisition of startup Pathway Medical for $63 million. Pathway has built an AI-powered clinical reference tool that doctors can use to ask questions about guidelines, drugs and trials. Pathway's answers are synthesized from medical literature, and Doximity said the Montreal-based startup has one of the largest structured d

Lightweight LSAT

Welcome to the lightweight LSAT The lightweight LSAT is a simple, proven, and completely free guide to the Law School Admissions Test. Who is this guide for? The lightweight LSAT is designed for students who are frustrated with their current way of approaching the LSAT. It doesn't assume you have any knowledge of the LSAT, but it will be most useful for someone who already has some experience studying. Additionally, the lightweight LSAT is written for students who are ambitious, aiming for a

Google is bringing image and PDF uploads to AI Mode

Google is updating AI Mode on desktop this week with the ability to process images, so you can ask it detailed questions about the pictures like you already can on mobile. In the coming weeks, the company is also adding support for PDF uploads on desktop, which could help you digest lengthy course or work materials. You can ask AI Mode to summarize the documents for you and ask follow-up questions that it will then answer by cross-referencing the materials you uploaded with information available

Join Our Next Livestream: Inside Katie Drummond’s Viral Interview With Bryan Johnson

What does it mean to be healthy in 2025? Bryan Johnson, an entrepreneur and venture capitalist who’s well known for his extreme attempts to slow the aging process, thinks he knows the answer. Does Johnson really have the healthiest body on Earth, as he claims? Will he achieve immortality through AI? Recently, WIRED global editorial director Katie Drummond visited Johnson’s home in California to sit down with him for WIRED's special Beyond Wellness edition. This wide-ranging interview is a must-

Show HN: Learn LLMs LeetCode Style

TorchLeet is broken into two sets of questions: Question Set: A collection of PyTorch practice problems, ranging from basic to hard, designed to enhance your skills in deep learning and PyTorch. LLM Set: A new set of questions focused on understanding and implementing Large Language Models (LLMs) from scratch, including attention mechanisms, embeddings, and more. Note Avoid using GPT. Try to solve these problems on your own. The goal is to learn and understand PyTorch concepts deeply. Table o

Join Our Livestream: Inside the AI Copyright Battles

What's going on right now with the copyright battles over artificial intelligence? Many lawsuits regarding generative AI’s training materials were initially filed back in 2023, with decisions just now starting to trickle out. Whether it’s Midjourney generating videos of Disney characters, like Wall-E brandishing a gun, or an exit interview with a top AI lawyer as he left Meta, WIRED senior writer Kate Knibbs has been following this fight for years—and she’s ready to answer your questions. Bring

Grok 4 appears to seek Elon Musk’s views when answering controversial questions

Elon musk and the xAI logo. Vincent Feuray | Afp | Getty Images When xAI's Grok 4 chatbot was launched on Wednesday, users and media outlets quickly began pointing out examples of it consulting its owner Elon Musk's views on controversial matters. CNBC was able to confirm that when asked to take a stance on some potentially contentious questions, the chatbot said it was analyzing posts from Musk while generating its answers. When asked, "Who do you support in the Israel vs Palestine conflict? O

Ars Live recap: Climate science in a rapidly changing world

The conversation then moved to the record we have of the Earth's surface temperatures and the role of Berkeley Earth in providing an alternate method of calculating those. While the temperature records were somewhat controversial in the past, those arguments have largely settled down, and Berkeley Earth played a major role in helping to show that the temperature records have been reliable. Lately, those temperatures have been unusually high, crossing 1.5° C above pre-industrial conditions for t

Perplexity launches AI-powered web browser for select group of subscribers

Perplexity AI on Wednesday launched a new artificial intelligence-powered web browser called Comet in the startup's latest effort to compete in the consumer internet market against companies like Google and Microsoft . Comet will allow users to connect with enterprise applications like Slack and ask complex questions via voice and text, according to a brief demo video Perplexity released on Wednesday. The browser is available to Perplexity Max subscribers, and the company said invite-only acce

Show HN: Dev atrophy test – Can you still code without AI?

Hey HN, I'm Per from Scrimba (YC S20), the code-learning platform. There's been a lot of talk lately about whether AI tools are causing skill atrophy amongst developers. We get a front-row seat to this, and we see more and more students struggle with basic concepts, and building apps on their own. This is almost always a consequence of relying too much on ChatGPT and vibe coding tools. So we built a small side project: https://devatrophy.com It's a test of your core web dev knowledge — no ha

Livestream Replay: Beginner Advice for Claude, a ChatGPT Alternative

Hello WIRED subscribers! Thank you to everyone who attended our most recent AI Unlocked livestream Q&A session, Chatbot Basics: Beginner Advice For Claude, a ChatGPT Alternative. Staff writer Reece Rogers and senior correspondent Kylie Robison provided an overview of Anthropic’s Claude chatbot, one of the most-used alternatives to OpenAI’s ChatGPT and popular with AI insiders. They also answered audience questions about all kinds of topics, such as the main differences between Claude and ChatGPT

Evaluating Long-Context Question and Answer Systems

While evaluating Q&A systems is straightforward with short paragraphs, complexity increases as documents grow larger. For example, technical documentation, novels and movies, as well as multi-document scenarios. Although some of these evaluation challenges also appear in shorter contexts, long-context evaluation amplifies issues such as: Information overload: Irrelevant details in large documents obscure relevant facts, making it harder for retrievers and models to locate the right evidence for

Show HN: I Built AskMedically – Get Research-Backed Answers to Medical Queries

Hi HN, I’ve built AskMedically – an AI-powered assistant that answers health and medical questions using real research papers from trusted medical sources like PubMed, Cochrane, etc. Whether you’re a healthcare enthusiast, patient, student, or professional – AskMedically helps you explore trusted medical knowledge without needing a medical degree or slogging through dozens of PDFs. Examples: • “Does intermittent fasting improve insulin sensitivity?” • “What are the benefits of creatine for

A Chinese firm has just launched a constantly changing set of AI benchmarks

Development of the benchmark at HongShan began in 2022, following ChatGPT’s breakout success, as an internal tool for assessing which models are worth investing in. Since then, led by partner Gong Yuan, the team has steadily expanded the system, bringing in outside researchers and professionals to help refine it. As the project grew more sophisticated, they decided to release it to the public. Xbench approached the problem with two different systems. One is similar to traditional benchmarking:

Think of a Number

My feed was recently clogged up with news articles reporting that Sam Altman thinks that AGI is here, or will be here next year, or whatever. I will refrain from giving even more air to this nonsense by linking to the stories. This kind of irresponsible hype-generation drives me nuts (although it also drives up stock prices so I can see why the tech bros are motivated to do it). Sure AI can have a good crack at undergraduate mathematics right now, and sure that’s pretty amazing. But our universi

Chemical knowledge and reasoning of large language models vs. chemist expertise

Benchmark corpus To compile our benchmark corpus, we utilized a broad list of sources (Methods), ranging from completely novel, manually crafted questions over university exams to semi-automatically generated questions based on curated subsets of data in chemical databases. For quality assurance, all questions have been reviewed by at least two scientists in addition to the original curator and automated checks. Importantly, our large pool of questions encompasses a wide range of topics and que