Latest Tech News

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

Filtered by: model Clear Filter

GPT-5 was supposed to simplify ChatGPT but now it has 4 new modes - here's why

Lance Whitney / Elyse Betters Picaro / ZDNET ZDNET's key takeaways GPT-5 now offers four different modes. You can still choose Auto to let ChatGPT determine how to respond. New choices add complexity to a model that was supposed to be simple. With the recent release of GPT-5, OpenAI touted its new model as a more efficient one that decides on its own how best to respond to your queries. Gone were the array of different modes that forced you to direct the AI based on whether you wanted a qui

Apple Watch Ultra 3 could remedy this current ‘Ultra’ discrepancy

Apple Watch Ultra 3 is coming very soon, and rumors indicate it could remedy a discrepancy that arose last year, when the Apple Watch Series 10 debuted several exclusive features that bested the latest Ultra model. Apple Watch Series 10 surpassed the latest Ultra model in several ways last year Last year after the Apple Watch Series 10 launched, I wrote an article that proved controversial. It was titled, “When Ultra doesn’t mean ‘best’ anymore.” The inspiration was Apple Watch Series 10 gain

Sam Altman Calls His Own AI Model "Annoying" After Being Forced to Raise It From the Dead

As OpenAI continues to try to tamp down on the chaos of GPT-5's disastrous release, CEO Sam Altman is throwing digs at the company's now-deprecated and controversial model. Late last week, OpenAI finally rolled out GPT-5, the latest version of its flagship large language model. Without warning, the company replaced all previous versions with the latest model — which, in addition to being deeply underwhelming, proved to have a colder, less obsequious tone than its predecessor, GPT-4o. The respo

Topics: 4o gpt model openai users

GPT-5 Doesn't Dislike You—It Might Just Need a Benchmark for Emotional Intelligence

Since the all-new ChatGPT launched on Thursday, some users have mourned the disappearance of a peppy and encouraging personality in favor of a colder, more businesslike one (a move seemingly designed to reduce unhealthy user behavior.) The backlash shows the challenge of building artificial intelligence systems that exhibit anything like real emotional intelligence. Researchers at MIT have proposed a new kind of AI benchmark to measure how AI systems can manipulate and influence their users—in

Topics: ai mit model new users

Ai2’s MolmoAct model ‘thinks in 3D’ to challenge Nvidia and Google in robotics AI

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 Physical AI, where robotics and foundation models come together, is fast becoming a growing space with companies like Nvidia, Google and Meta releasing research and experimenting in melding large language models (LLMs) with robots. New research from the Allen Institute for AI (Ai2) aims to challenge Nvidia and Google in physical AI with th

GPT-5 was meant to cut choices, but OpenAI just added multiple modes - why?

Lance Whitney / Elyse Betters Picaro / ZDNET ZDNET's key takeaways GPT-5 now offers four different modes. You can still choose Auto to let ChatGPT determine how to respond. New choices add complexity to a model that was supposed to be simple. With the recent release of GPT-5, OpenAI touted its new model as a more efficient one that decides on its own how best to respond to your queries. Gone were the array of different modes that forced you to direct the AI based on whether you wanted a qui

OpenAI brings back GPT-4o after user revolt

On Tuesday, OpenAI CEO Sam Altman announced that GPT-4o has returned to ChatGPT following intense user backlash over its removal during last week's GPT-5 launch. The AI model now appears in the model picker for all paid ChatGPT users by default (including ChatGPT Plus accounts), marking a swift reversal after thousands of users complained about losing access to their preferred models. The return of GPT-4o comes after what Altman described as OpenAI underestimating "how much some of the things t

OpenAI brings GPT-4o back as a default for all paying ChatGPT users, Altman promises ‘plenty of notice’ if it leaves again

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 OpenAI is once again making GPT-4o — the large language model (LLM) that powered ChatGPT before last week’s launch of GPT-5 — a default option for all paying users, that is, those who subscribe to the ChatGPT Plus ($20 per month), Pro ($200 per month), Team ($30 per month), Enterprise, or Edu tiers, no longer requiring users to toggle on a

Evaluating GPT5's reasoning ability using the Only Connect game show

Given the proliferation of reasoning models, we wanted to go beyond knowledge-based benchmarks to test reasoning abilities such as pattern recognition, lateral thinking, abstraction, contextual reasoning (accounting for British cultural references), and multi-step inference. In addition to reasoning, we aimed to assess how effectively models make decisions when presented with judgment calls—such as choosing between making an educated guess based on available clues or calling a function to retri

ChatGPT won’t remove old models without warning after GPT-5 backlash

Posts from this author will be added to your daily email digest and your homepage feed. After the backlash to replacing its 4o model with GPT-5, OpenAI will no longer get rid of old models without a heads up. “In retrospect, not continuing to offer 4o, at least in the interim, was a miss,” Nick Turley, OpenAI’s head of ChatGPT, said on Tuesday. In an interview with The Verge, he said it was surprising to see the “level of attachment” people had to 4o. “It’s not just change that is difficult fo

Topics: 4o gpt model said users

OpenAI adds new GPT-5 models, restores o3, o4-mini and it's a mess all over again

One of the few things many disliked about ChatGPT was the confusing number of models. OpenAI claimed GPT-5 would fix this, but it seems to have made it worse. A new update is rolling out to ChatGPT. It doesn't upgrade GPT-5, but instead adds more options that some of you would love. Previously, GPT-5 had two variants - GPT (auto-rotates between reasoning and non-reasoning) and GPT-Thinking (reasoning). GPT-5 model selector Source: BleepingComputer Today's update populates the ChatGPT 5 mode

Evaluating LLMs playing text adventures

What we’ll do is set a low-ish turn limit and see how much they manage to accomplish in that time.1 Another alternative for more linear games is running them multiple times with a turn limit and seeing how often they get past a particular point within that turn limit. Given how much freedom is offered to players of text adventures, this is a difficult test. It’s normal even for a skilled human player to immerse themselves in their surrounding rather than make constant progress. I wouldn’t be su

ChatGPT’s model picker is back, and it’s complicated

When OpenAI launched GPT-5 last week, the company said the model would simplify the ChatGPT experience. OpenAI hoped GPT-5 would act as a sort of “one size fits all” AI model with a router that would automatically decide how to best answer user questions. The company said this unified approach would eliminate the need for users to navigate its model picker — a long, complicated menu of AI options that OpenAI CEO Sam Altman has publicly said he hates. But it looks like GPT-5 is not the unified A

Topics: ai gpt model openai users

Liquid AI wants to give smartphones small, fast AI that can see with new LFM2-VL model

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 Liquid AI has released LFM2-VL, a new generation of vision-language foundation models designed for efficient deployment across a wide range of hardware — from smartphones and laptops to wearables and embedded systems. The models promise low-latency performance, strong accuracy, and flexibility for real-world applications. LFM2-VL builds o

Topics: ai lfm2 liquid models vl

What are Apple’s options for an AI acquisition beyond Perplexity?

Since Apple’s latest earnings call, talk of a potential Perplexity acquisition has quieted down (the fact that Perplexity was once again allegedly caught red-handed sidestepping content restrictions didn’t help). Meanwhile, with the ever-increasing number of engineers from its Foundation Models team jumping ship, Apple’s need for fresh talent is getting more urgent by the day. But if Perplexity is a no-go, who else could Apple buy? I used to agree with Jason Snell’s frequent argument on the Up

OpenAI Scrambles to Update GPT-5 After Users Revolt

OpenAI’s GPT-5 model was meant to be a world-changing upgrade to its wildly popular and precocious chatbot. But for some users, last Thursday’s release felt more like a wrenching downgrade, with the new ChatGPT presenting a diluted personality and making surprisingly dumb mistakes. On Friday, OpenAI CEO Sam Altman took to X to say the company would keep the previous model, GPT-4o, running for Plus users. A new feature designed to seamlessly switch between models depending on the complexity of t

Launch HN: Design Arena (YC S25) – Head-to-head AI benchmark for aesthetics

Hi HN, I’m Grace from Design Arena ( https://www.designarena.ai/ ) - we’re building a crowdsourced benchmark for AI-generated visuals (websites, images, video, and more). We put AI models and builder tools in head-to-head comparisons that get voted on by real users from around the world. Think “Hot or Not” for the AI era :) (Btw, when we say real users we mean real users, so you may get a captcha on the site. Sorry, but we have to use every bot protection available! We only want human ratings,

Are Gesture-Enabled AirPod Live Translations Incoming? iOS 26 Beta Suggests Yes

Some models of Apple's popular AirPods may soon be able to do live, in-person language translations when you squeeze both stems at the same time. According to an image posted by websites including 9to5Mac, the touch gesture is featured in a system asset that's part of Apple iOS 26 developer beta 6. In the image, the gesture is shown on a pair of AirPods with words in languages including English, Spanish, German, French and Portuguese. In June, Apple showed off AI-powered live translations featu

Deals: 32GB M4 MacBook Air $200 off, Black/Natural Apple Watch Ultra 2 $150 off, AirPods 4 $99, more

Today’s 9to5Toys Lunch Break deals are now ready to roll starting with the M4 MacBook Air. Alongside entry-level models from $799, we are also still tracking rare $200 price drops on a 24GB model with 1TB of storage an heavily upgraded variant with 32GB of RAM today. Moving over to Apple Watch Ultra 2 – we have both the Natural and Black Titanium models at $150 off the list price as well as ongoing deals on AirPods 4, Apple chargers, iPad A16, and more. Scope it all out down below. Rare price d

Nexus: An Open-Source AI Router for Governance, Control and Observability

Today, we're excited to introduce Nexus - a powerful AI router designed to optimize how AI agents interact with multiple MCP tools and Large Language Models. Nexus serves as a central hub that aggregates Model Context Protocol (MCP) servers while providing intelligent LLM routing, security and governance capabilities. Nexus is an AI router that solves two critical challenges in the AI ecosystem: MCP Server Aggregation: Instead of managing connections to multiple MCP servers individually, Nexus

Evaluating LLMs Playing Text Adventures

What we’ll do is set a low-ish turn limit and see how much they manage to accomplish in that time.1 Another alternative for more linear games is running them multiple times with a turn limit and seeing how often they get past a particular point within that turn limit. Given how much freedom is offered to players of text adventures, this is a difficult test. It’s normal even for a skilled human player to immerse themselves in their surrounding rather than make constant progress. I wouldn’t be su

This Gemini UI change should’ve been the default from the start (APK teardown)

Edgar Cervantes / Android Authority TL;DR Google could move Gemini’s AI model switcher to the bottom of your smartphone screen. This would make it easier to switch AI models with one hand compared to the current UI. Google could also bring a UI tweak to Gemini’s video menu. Google’s Gemini chatbot lets you choose between several AI models for your specific needs. You can use the flash models if you value quick, responsive answers, or the Pro models if you need more in-depth answers. Now, it

The GPT-5 rollout has been a big mess

It's been less than a week since the launch of OpenAI's new GPT-5 AI model, and the rollout hasn't been a smooth one. So far, the release sparked one of the most intense user revolts in ChatGPT's history, forcing CEO Sam Altman to make an unusual public apology and reverse key decisions. At the heart of the controversy has been OpenAI's decision to automatically remove access to all previous AI models in ChatGPT (approximately nine, depending on how you count them) when GPT-5 rolled out to user

OpenAI is editing its GPT-5 rollout on the fly — here’s what’s changing in ChatGPT

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 OpenAI’s launch of its most advanced AI model GPT-5 last week has been a stress test for the world’s most popular chatbot platform with 700 million weekly active users — and so far, OpenAI is openly struggling to keep users happy and its service running smoothly. The new flagship model GPT-5 — available in four variants of different speed

AI summaries can downplay medical issues for female patients, UK research finds

The latest example of bias permeating artificial intelligence comes from the medical field. A new study surveyed real case notes from 617 adult social care workers in the UK and found that when large language models summarized the notes, they were more likely to omit language such as "disabled," "unable" or "complex" when the patient was tagged as female, which could lead to women receiving insufficient or inaccurate medical care. Research led by the London School of Economics and Political Sci

OpenAI is testing 3,000-per-week limit for GPT-5 Thinking

OpenAI has responded to criticism that it shipped GPT-5 with token limits to minimize cost and maximize profit not with words, but rather with a new 3,000-per-week limit. In a series of posts on X, Sam Altman confirmed that OpenAI is working on a 3,000-per-week limit for GPT-5 Thinking messages for Plus users. This will increase the reasoning rate limits available today, but OpenAI does not plan to stop at just this. Sam Altman claims that OpenAI will soon raise all model-class rate limits "a

Here are all the GPT-5 updates OpenAI has rolled out since launch

SOPA Images/Contributor/Getty ZDNET's key takeaways: OpenAI released its long-awaited GPT-5 on Thursday. Some users complained GPT-5 was inferior to its predecessor, 4o. In response, the company announced a flurry of changes. OpenAI released GPT-5, the long-awaited upgrade to the model which powers ChatGPT, Thursday. In typical OpenAI fashion, the release has included plenty of twists, turns, and drama. It was almost inevitable that the new model would disappoint a significant number of pe

Users Were So Addicted to GPT-4o That They Immediately Cajoled OpenAI Into Bringing It Back After It Got Killed

Last week, OpenAI startled the world by announcing that its long-awaited GPT-5 would replace all of its previous models, The move sparked outrage. Apart from being severely underwhelmed by the performance of OpenAI's newest offering, power users immediately started to beg CEO Sam Altman to bring back preceding models, often for a reason that had little to do with intelligence, artificial or otherwise: they were attached to it on an emotional level. "Why are we getting rid of the variants and 4

Token growth indicates future AI spend per dev

Kilo just broke through the 1 trillion tokens a month barrier on OpenRouter for the first time. Each of the open source family of AI coding tools (Cline, Roo, Kilo) is growing rapidly this month. Part of this growth is caused by Cursor and Claude starting to throttle their users. We wrote about Cursor at the beginning of July and about Claude in the second half of July. Their throttling sent users to the open source family of AI coding tools causing the increases you see in the graphs above. C

Deals: Apple Watch Series 10 new low up to $150 off, M4 Pro MacBook Pro $299 off, iPad Air, more

Today’s 9to5Toys Lunch Break deals are kicking off with the lowest prices we have tracked online for GPS + Cell Apple Watch Series 10 models. Alongside GPS only variants at $100 off, you’ll now find the cell variants at up to $149 off in brand new condition with a full Apple warranty in tow. Those deals also join one of the best prices to date on the M4 Pro MacBook Pro at $299 off the list price, ongoing all-time lows on M3 iPad Air, and more. Everything awaits below. Apple Watch Series 10 Cell