Technologies

AI Gets Smarter, Safer, More Visual With GPT-4 Update, OpenAI Says

If you subscribe to ChatGPT Plus, you can try it out now.

The hottest AI technology foundation got a big upgrade Tuesday with OpenAI’s GPT-4 release now available in the premium version of the ChatGPT chatbot.

GPT-4 can generate much longer strings of text and respond when people feed it images, and it’s designed to do a better job avoiding artificial intelligence pitfalls visible in the earlier GPT-3.5, OpenAI said Tuesday. For example, when taking bar exams that attorneys must pass to practice law, GPT-4 ranks in the top 10% of scores compared with the bottom 10% for GPT-3.5, the AI research company said.

GPT stands for Generative Pretrained Transformer, a reference to the fact that it can generate text on its own — now up to 25,000 words with GPT-4 — and that it uses an AI technology called transformers that Google pioneered. It’s a type of AI called a large language model, or LLM, that’s trained on vast swaths of data harvested from the internet, learning mathematically to spot patterns and reproduce styles. Human overseers rate results to steer GPT in the right direction, and GPT-4 has more of this feedback.

OpenAI has made GPT available to developers for years, but ChatGPT, which debuted in November, offered an easy interface ordinary folks can use. That yielded an explosion of interest, experimentation and worry about the downsides of the technology. It can do everything from generating programming code and answering exam questions to writing poetry and supplying basic facts. It’s remarkable if not always reliable.

ChatGPT is free, but it can falter when demand is high. In January, OpenAI began offering ChatGPT Plus for $20 per month with assured availability and, now, the GPT-4 foundation. Developers can sign up on a waiting list to get their own access to GPT-4.

GPT-4 advancements

«In a casual conversation, the distinction between GPT-3.5 and GPT-4 can be subtle. The difference comes out when the complexity of the task reaches a sufficient threshold,» OpenAI said. «GPT-4 is more reliable, creative and able to handle much more nuanced instructions than GPT-3.5.»

Another major advance in GPT-4 is the ability to accept input data that includes text and photos. OpenAI’s example is asking the chatbot to explain a joke showing a bulky decades-old computer cable plugged into a modern iPhone’s tiny Lightning port. This feature also helps GPT take tests that aren’t just textual, but it isn’t yet available in ChatGPT Plus.

Another is better performance avoiding AI problems like hallucinations — incorrectly fabricated responses, often offered with just as much seeming authority as answers the AI gets right. GPT-4 also is better at thwarting attempts to get it to say the wrong thing: «GPT-4 scores 40% higher than our latest GPT-3.5 on our internal adversarial factuality evaluations,» OpenAI said.

GPT-4 also adds new «steerability» options. Users of large language models today often must engage in elaborate «prompt engineering,» learning how to embed specific cues in their prompts to get the right sort of responses. GPT-4 adds a system command option that lets users set a specific tone or style, for example programming code or a Socratic tutor: «You are a tutor that always responds in the Socratic style. You never give the student the answer, but always try to ask just the right question to help them learn to think for themselves.»

«Stochastic parrots» and other problems

OpenAI acknowledges significant shortcomings that persist with GPT-4, though it also touts progress avoiding them.

«It can sometimes make simple reasoning errors … or be overly gullible in accepting obvious false statements from a user. And sometimes it can fail at hard problems the same way humans do, such as introducing security vulnerabilities into code it produces,» OpenAI said. In addition, «GPT-4 can also be confidently wrong in its predictions, not taking care to double-check work when it’s likely to make a mistake.»

Large language models can deliver impressive results, seeming to understand huge amounts of subject matter and to converse in human-sounding if somewhat stilted language. Fundamentally, though, LLM AIs don’t really know anything. They’re just able to string words together in statistically very refined ways.

This statistical but fundamentally somewhat hollow approach to knowledge led researchers, including former Google AI researchers Emily Bender and Timnit Gebru, to warn of the «dangers of stochastic parrots» that come with large language models. Language model AIs tend to encode biases, stereotypes and negative sentiment present in training data, and researchers and other people using these models tend «to mistake … performance gains for actual natural language understanding.»

OpenAI Chief Executive Sam Altman acknowledges problems, but he’s pleased overall with the progress shown with GPT-4. «It is more creative than previous models, it hallucinates significantly less, and it is less biased. It can pass a bar exam and score a 5 on several AP exams,» Altman tweeted Tuesday.

One worry about AI is that students will use it to cheat, for example when answering essay questions. It’s a real risk, though some educators actively embrace LLMs as a tool, like search engines and Wikipedia. Plagiarism detection companies are adapting to AI by training their own detection models. One such company, Crossplag, said Wednesday that after testing about 50 documents that GPT-4 generated, «our accuracy rate was above 98.5%.»

OpenAI, Microsoft and Nvidia partnership

OpenAI got a big boost when Microsoft said in February it’s using GPT technology in its Bing search engine, including a chat features similar to ChatGPT. On Tuesday, Microsoft said it’s using GPT-4 for the Bing work. Together, OpenAI and Microsoft pose a major search threat to Google, but Google has its own large language model technology too, including a chatbot called Bard that Google is testing privately.

Also on Tuesday, Google announced it’ll begin limited testing of its own AI technology to boost writing Gmail emails and Google Docs word processing documents. «With your collaborative AI partner you can continue to refine and edit, getting more suggestions as needed,» Google said.

That phrasing mirrors Microsoft’s «co-pilot» positioning of AI technology. Calling it an aid to human-led work is a common stance, given the problems of the technology and the necessity for careful human oversight.

Microsoft uses GPT technology both to evaluate the searches people type into Bing and, in some cases, to offer more elaborate, conversational responses. The results can be much more informative than those of earlier search engines, but the more conversational interface that can be invoked as an option has had problems that make it look unhinged.

To train GPT, OpenAI used Microsoft’s Azure cloud computing service, including thousands of Nvidia’s A100 graphics processing units, or GPUs, yoked together. Azure now can use Nvidia’s new H100 processors, which include specific circuitry to accelerate AI transformer calculations.

AI chatbots everywhere

Another large language model developer, Anthropic, also unveiled an AI chatbot called Claude on Tuesday. The company, which counts Google as an investor, opened a waiting list for Claude.

«Claude is capable of a wide variety of conversational and text processing tasks while maintaining a high degree of reliability and predictability,» Anthropic said in a blog post. «Claude can help with use cases including summarization, search, creative and collaborative writing, Q&A, coding and more.»

It’s one of a growing crowd. Chinese search and tech giant Baidu is working on a chatbot called Ernie Bot. Meta, parent of Facebook and Instagram, consolidated its AI operations into a bigger team and plans to build more generative AI into its products. Even Snapchat is getting in on the game with a GPT-based chatbot called My AI.

Expect more refinements in the future.

«We have had the initial training of GPT-4 done for quite awhile, but it’s taken us a long time and a lot of work to feel ready to release it,» Altman tweeted. «We hope you enjoy it and we really appreciate feedback on its shortcomings.»

Editors’ note: CNET is using an AI engine to create some personalfinance explainers that are edited and fact-checked by our editors. Formore, see this post.

Technologies

My Kid Wanted Video Games. I Was Against It. This Console Gave Us Both the Win

The movement-based Nex Playground might be the antidote to parental screen time guilt.

When our 8-year-old started asking for video games, I knew we were about to engage in an uphill battle. Anytime we’ve been to friends’ houses with gaming consoles, he goes full zombie mode, then has an epic meltdown once the sensory overload wears off. And since he inevitably ropes his 6-year-old brother in, we’re essentially sealing both their fates.

So when our neighbors started raving about a movement-based gaming console called Nex Playground, my first instinct was to shut it down. The words «gaming console» alone were enough to put me in a mental block. Add in my own memories of Wii tennis sessions where I nearly took out the ceiling fan, and I was firmly in the «no» camp.

But after doing a little more research, I was intrigued enough to try it out.

Screen time isn’t something I take lightly. With three kids ages 2 to 8, my husband and I have always been intentional about how and what they watch. They don’t have their own tablets, and most of their screen time happens on our family TV, which means whatever the oldest is exposed to quickly trickles down to our toddler. So anything we bring into the house has to work for all of them. Tall order, I know, but the Nex Playground gets surprisingly close.

Getting started is easy

The console itself is refreshingly simple. It’s a small cube, slightly larger than a Rubik’s cube, with a circular camera and motion sensor, a light indicator and two ports for power, and an HDMI connection to the TV. There’s no controller beyond a basic remote for navigating menus. For most games, your body is the controller.

Setup is quick. Plug it in, connect it to your TV, and you’re ready to go. It doesn’t store video or upload footage to the cloud, which was an immediate plus. It also comes with a magnetic privacy cover that you can put on the lens when it’s not in use.

At $250, it’s not cheap, but it’s less than some of the popular gaming consoles for this age range, like the Nintendo Switch 2. That gets you a five-game starter pack: Fruit Ninja, Go Keeper (soccer), Starri (think Guitar Hero for your whole body), Party Fowl (an AR emoji frenzy) and Whack-a-Mole. Additional games require a subscription: $89 a year or $49 for three months, which unlocks a library of 50-plus games and counting. New titles dropped even as I was writing this.

The library spans a surprisingly wide range. There are board game adaptations like Connect Four and Candy Land, character-driven games with Peppa Pig, Bluey and the Ninja Turtles, and sports like baseball and, yes, tennis — minus the ceiling fan hazard. There’s even parent-friendly content like Zumba workouts, which I may or may not have fully committed to on a rainy afternoon.

Even my toddler has gotten in on the action, mostly bouncing her way through Hungry Hungry Hippos when her brothers finally concede.

Gameplay is where it wins

The movements range from swinging your arms to keep a ball in motion, hopping or full-body launches that are far more aggressive than what the game actually requires. (I’m not about to tell the kids otherwise.) After a 45-minute session, my kids are tired and sometimes even drenched in sweat. The Nex Playground entertains and burns energy in one fell swoop.

The graphics also seem intentionally simple and arcade-like, which fits the minimalist play experience. There’s no POV storyline to get lost in, no leveling up into a new world at 9 p.m. on a school night. Some games keep score, which awakens my kids’ competitive streak, but the vibe is more collaborative and hasn’t been the catalyst for more fighting like other games. If anything, it’s done the opposite.

I still don’t love defaulting to a screen when my kids are bored, so we try to use it in moderation. In our house, piano practice is the only thing that unlocks weekend play time, and the fact that they’ll sit at the piano for a full hour tells you everything you need to know.

The verdict that matters most

But the real test: Does it hold up to an 8-year-old who was dead set on a Nintendo Switch?

Short answer: yes. At least for now. He’d still pick the Switch if you asked him, but not for the reasons you’d expect.

«The Playground is more tiring,» he told me, which only helped seal the deal for me. His current favorite is Homerun Hitters. «It’s basically a baseball game where you go against ranked global players. Me and my brother are really good at it.»

This from a kid whose primary hobby is annoying his younger brother. The fact that he said «me and my brother» as a collective was an unexpected bonus.

The Switch may still show up on the Christmas list this year. And realistically, I know I’m on borrowed time. As kids get older, «cool» becomes the currency, and a motion-based cube probably won’t hold up against an Xbox or a Switch once playdates turn into side-by-side gaming sessions.

The Nex Playground isn’t a replacement for those. It’s more of a detour; it gives them a taste of gaming without all the usual side effects. Even if I do eventually cave, I can still see it sticking around for the occasional family game night or as a rainy-day sibling diffuser.

In the meantime, I’ll relish this simpler version of gaming while I still can. He’s not exactly rushing me to return this review unit. More importantly, neither am I.

Technologies

Don’t Wait for New Emoji in iOS 26.4, Here’s How to Create Them on Your Own

If your iPhone has Apple Intelligence, you can create your own emoji now.

Apple will likely add new emoji to your iPhone when the company releases iOS 26.4. Those new emoji could include an orca, a distorted smiley face and more. According to Emojipedia, there are 3,953 emoji with more on the way. The current list of emoji include smileys, sports players, weather conditions and flags. But there’s no emoji for a dog wearing pajamas, a plate with burgers and fries and many other things. But if you have Genmoji on your iPhone you can create these emoji and many more.

Apple released iOS 18.2 in 2024 and the company introduced its own emoji generator, called Genmoji, to Apple Intelligence-capable iPhones at that time. The Unicode Standard, a universal character encoding standard, is responsible for creating new emoji, and approved emoji are added to all devices once a year. With Genmoji, you don’t have to wait for new emoji to appear on your iPhone each year. You can just create them as you need them.

Read on to learn how to use Genmoji on iPhone to create your own custom emoji. Just note that only iPhones with Apple Intelligence, like the iPhone 17 lineup, can use Genmoji at this time.

How to make custom emoji

1. Open Messages and go into a chat.
2. Tap the plus (+) button next to your text box.
3. Tap Genmoji.

You can then type a description of an emoji into the text box near the bottom of your screen and tap the check mark on your keyboard to enter that description into Genmoji. You can also tap different suggestions and themes that are right above the text box. And with iOS 26 or later, you can also combine and use emoji to create others rather than describing a new emoji or using suggestions.

Your iPhone will generate a series of new emoji for you to pick from according to your description, and you can swipe through these new emoji. When you find the one you want, tap Add in the top right corner of your screen and the new emoji will be available to use as an emoji, tapback or a sticker. Now you don’t have to wait for the Unicode Standard to propose, create and bring new emoji to devices.

For more iOS news, here’s what to know about iOS 26.3.1 and iOS 26.3. You can also check out our iOS 26 cheat sheet for other tips and tricks.

Technologies

Apple Needs to Launch Its Foldable iPhone Flip in 2026. Here’s Why

Commentary: Foldables are everywhere now and Apple is the only major phone-maker without one.

I love Apple’s flagship cosmic orange iPhone 17 Pro — even when I managed to turn mine pink — but I was disappointed not to see the company’s long-rumored foldable iPhone Flip. Pretty much every major Android phone-maker, including Samsung, Google, Motorola, OnePlus, Xiaomi and Honor are now multiple generations into their own folding phone lineups, with the hardware continuing to become more and more refined with each revision. Oppo is now in its fifth year of foldables and its latest Find N6 is the result of those years of development. Apple isn’t even at step one yet and it’s beginning to feel like it’s late to the party. That might be a problem.

Apple dominates in the premium phone category, but foldables — which fit into the premium space in terms of price — are already nipping at its heels, with Motorola telling CNET that 20% of customers buying its Razr foldable jumped ship from Apple. Meanwhile, Samsung is in the seventh generation of its Flip and Fold series. As Lisa Eadicicco discovered during a visit to Seoul, «foldables are everywhere» in Samsung’s home country of South Korea.

With nearly every major Android phone-maker entering the foldable market, Apple risks losing potential customers. It also runs the risk of letting a rival like Samsung or Motorola becoming the go-to name for foldables, which could make it harder for Apple to make an impact if it eventually launches its own device. Furthermore, early adopters drawn to foldable tech may be too entrenched in the Android ecosystem by the time Apple’s phone arrives to want to switch to iOS.

Apple is unlikely to be worried. It’s estimated that around 20 million foldables from all manufacturers were sold worldwide in 2023, while Apple reportedly sold 26.5 million iPhone 14 Pro Max handsets in the first half of that year alone. In 2024, foldable sales were flat — and 2025 didn’t fare much better, according to analysts at CounterPoint Research, although Samsung did report record numbers of preorders for its most recent foldable. Clearly, Apple feels it has yet to miss the boat.

Apple has always found success in biding its time, observing the industry and launching its own take on a product when it’s ready. Apple didn’t invent phones, tablets, smartwatches or computers, but it found ways to take existing products and make them more useful, more valuable in day-to-day life and — dare I say — more exciting. It’s why the iPhone, iPad, Apple Watch and Mac lines dominate the market today.

For me, I need to see Apple’s take on the foldable phone. I’ve written before about how disappointed I am in foldables. I’ve been a mobile reporter for over 14 years and phones have become increasingly dull as they’ve converged to become slight variations on the same rectangular slab.

Read more: Best Flip Phone for 2026

Foldables promised something new, something innovative, something that briefly sparked some excitement in me, but years in, that excitement has dwindled to the point of being extinguished. They are fine products and while I like the novelty of a screen that bends, they’re not a revolution in how we interact with our phones. Not in the way that the arrival of the touchscreen was when we were still pushing buttons to type out texts.

I did hope that Google’s Pixel Fold would be the phone to catapult the foldable forward, and while the recent Pixel 10 Pro Fold — the second generation of Google’s foldable — does offer some great updates, it still doesn’t offer any kind of revolution. Instead, it feels more like a «me too» move from Google. Ditto for the OnePlus Open. So I’m left instead to look toward Apple, a company with a track record for product revolutions, to create a new take on the genre that genuinely drives forward how we use our phones.

That innovation won’t just come from the product design. Apple works closely with its third-party software developers, and it’s that input that would help a folding iPhone become genuinely useful. My biggest complaint around foldables right now is that while the hardware is decent, the devices are essentially just running standard versions of Android with a handful of UI tweaks thrown in. They’re just regular phones that just happen to bend.

Few Android developers are embracing the folding format, and it’s not difficult to see why; the users aren’t there in sufficient numbers yet to justify the time and expense to adapt their software across a variety of screen sizes. The multiple folding formats already available mean Android foldables face the same fragmentation issue that has plagued the platform since the beginning. Android-based foldables are simply a more difficult platform for developers to build for than regular phones. Apple would be able to change that, as it proved with the iPhone and iPad.

Given Apple’s close relationships with top-tier developers — not to mention its own vast developer team — I expect an eventual Apple foldable to offer innovations that make it more than just an iPhone that folds in half.

And I truly hope it does. I want to look forward to tech launches again. I want to feel excited to get a new gadget in my hands and feel that «wow» moment as I do something transformative for the first time.

In short, I don’t want to be bored by technology anymore. Apple, it’s over to you.