Connect with us

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

Onimusha Way of the Sword Hands-On: Back to the Demon-Killing Samurai Grind

Capcom’s triumphant return to the Onimusha franchise felt pretty rote in a short demo — until the outstanding boss fight.

It’s been 19 years since Capcom released another game progressing the story of its samurai action horror franchise, Onimusha — but in 2026, Onimusha: Way of the Sword will arrive as the next game in the series. Ahead of Gamescom, Capcom gave US media a preview of the game with a short 15-minute demo, a mere taste to show us where the next game is going in the two decades since the last mainline Onimusha.

Onimusha: Way of the Sword continues the franchise’s tradition of basing characters on historic figures and then having them fight demons. Players take on the role of Miyamoto Musashi — one of the most famous swordsmen in Japanese history — tasked with destroying the Genma, an army of demons plaguing the land. You’ll hack, you’ll slash, you’ll parry, you’ll kick ass.

The Onimusha games have always been something of an arcade hack-and-slash with parrying mechanics, a novelty when the series debuted in the early 2000s, but which other games have adopted in the decades since. Period samurai games have been in vogue in recent years, from 2019’s Sekiro: Shadows Die Twice to 2020’s Ghost of Tsushima (and Ghost of Yotei releasing soon) to this year’s Assassin’s Creed: Shadows. 

While my demo with Onimusha: Way of the Sword was brief, it seems like Capcom has resisted mimicking the stealth and hyper-mobility of those more modern samurai games and preserved the idiosyncrasies of the originals. Musashi receives the power of the Oni gauntlet, through which he can absorb orbs of the hordes of demons he slays — and no, you still can’t jump.

Most of the modernizations to Onimusha: Way of the Sword is through the parry system, which has four different ways to deflect enemy attacks, presuming you time it correctly. Yet it’s still a game about cutting apart demons with your sword in frankly brutal fashion, depleting their block meter to get in a final slash that will frequently slice them in half in a gruesomely satisfying fashion. 

A short slice of Onimusha: Way of the Sword

The demo opened up with the player as Musashi walking down a forest path to get to a temple, passing fleeing villagers and sword-wielding demon soldiers pursuing them. Killing them was easy — the game was on the Action difficulty (with an even easier Story difficulty if I wanted), and even with my middling Sekiro and Elden Ring skills, it was a breeze, so I’m hoping for a tougher option when the game releases.

A dark fog gathered around the temple, which Musashi called Malice, which sounds like a concentration of demonic presence (but what do I know). As I approach the temple gates, the Oni gauntlet, the soul-gathering demon armor piece that’s iconic to the franchise, speaks to Musashi. When he moves to touch a glowing orb, ghostly memories of villagers march into the temple. To enter myself, I had to use Oni Vision — basically a Batman-style detective sight — to find the right spiritual binding to cut.

Naturally, more demon soldiers await me in the temple, giving me a chance to use my special weapons — a pair of twin blades I can summon when I’ve revved up enough energy in the Oni Power Gauge using my regular sword. Presumably, I’ll get to use a variety of magical arms this way in the full game, but it does seem like most of the combat will be using my trusty katana (and a lot of parrying). 

I finally entered the main sanctuary of the temple overlooking the valley below, which had another ghostly memory for me to watch — recalling the corrupting Malice convincing elderly villagers to toss their children into the abyss. Grim stuff. But I was quickly shaken out of my reverie by an old friend of Musashi’s who greeted him with a blade. There’s some unspoken history between your legendary swordsman and the seemingly unhinged newcomer (named Sasaki Ganryu), but he’s got an Oni Gauntlet too, and a desire to cut you down.

Unlike the fodder I’d fought before, Ganryu as a boss was a satisfying and lengthy fight, requiring plenty of counters and measured attacks to break your opponent’s stance. When you do, you’re given a choice of where to land your critical hit — in the boss’s body for extra damage, or in his Oni Gauntlet to get more orb currency. 

While you have some healing items to use normally, I couldn’t apply them during the boss battle — but successive attacks will make healing orbs pop out, rewarding precise play while forgiving missteps. It’s a promising alternative to other punishing slash-and-parry games, like FromSoftware’s Souls titles and their imitators. 

With the boss defeated, the demo ended, and thus our first look at a brand-new Onimusha game in two decades. While Onimusha: Way of the Sword comes in the wake of several other action games set in historical Japan, Capcom’s contribution has the arcade feel and brutality, combined with demonic mystery, that could set it apart from more grounded and realistic samurai simulators.

Onimusha: Way of the Sword is coming out in 2026.

Continue Reading

Technologies

Why Wait for the Pixel 10 Pro Fold When You Can Grab the Pixel 9 Pro Fold for $700 Off?

This sleek and stunning large-screen foldable is just $1,099 at Best Buy — the all-time lowest price I’ve seen.

I don’t know about you, but I’m eagerly awaiting news of Google’s new Pixel 10 phones today at its Made by Google event, including the expected Pixel 10 Pro Fold. But while we all wait, there are still plenty of good deals to be had on what’s about to be the last-gen of Pixel phones. They’re still excellent devices too, so if you’re looking for a new phone while saving some money, this could be the best way to do that.

Best Buy has a deal available right now that’ll get you the excellent Pixel 9 Pro Fold for a new-low price of just $1,099. This is the lowest I’ve ever seen the device with no strings attached. If you are looking to sign up with a carrier too though, you can save an extra $100 on it by doing so as you buy it.

In our review of the Pixel 9 Pro Fold, mobile expert Lisa Eadicicco said the device is «more polished and practical» in terms of design than its predecessor, feeling like «a big leap» forward. As well as the improved physical form, she praised its larger screens and the substantial seven-year software update commitment Google has made for this model. One major issue was the price, but this massive $700 discount goes a long way toward addressing that.

Hey, did you know? CNET Deals texts are free, easy and save you money.

This 2024 foldable features a durable 6.3-inch Corning Gorilla Glass Victus 2 cover screen and a stunning 8-inch Super Actua Flex interior display. You can even run two apps side-by-side, making it great for multitasking. On the inside, the basic model comes with 256GB of storage and 16GB of RAM, and is equipped with Google’s advanced G4 Tensor chip. All of this in a phone that weighs in at just 9.1 ounces and is just 0.2 inches thick when open (or 0.4 inches when closed).

Why this deal matters

Google’s foldable phone is the most expensive in the lineup, meaning tech fans have to pay a premium to get one. With this deal, the price is cut substantially, dropping it down to a new record-low. The price reduction here makes this one of the best Google Pixel deals around for anyone not looking to trade an old phone in or get a new line. It’s a solid discount.

Just don’t forget that we’re expecting Google to release the Pixel 10 Pro Fold alongside its other devices at an event today.

Continue Reading

Technologies

Verizon’s Simple Mobile and Total Wireless Plans Expand International Features

If you frequently call or visit countries outside the US, these prepaid phone plan improvements could help your travel budget.

Travel patterns are changing, with more people in the US venturing outside the country and needing to stay connected. To accommodate this surge, Verizon is boosting the international options in its prepaid Simple Mobile and Total Wireless phone brands starting Aug. 28.

Most phone plans include options to call and text people in Canada and Mexico, along with some roaming options for US travelers. Depending on the plan, those options are extended to many other countries around the globe. But you often end up paying for a more expensive plan than you need in order to get the travel benefits.

Now, the Simple Mobile and Total Wireless brands are expanding their international options and increasing data allotments so you can travel using the plan you carry every day without the stress of finding local coverage.

See also: Make sure you have a good phone for traveling.

Simple Mobile changes

The Simple Mobile prepaid plans include an allotment of high-speed data before downshifting to slower unlimited data. Be sure to check the details for specific countries on the site. Here’s how they break down:

  • The $25 Unlimited plan includes 15GB of high-speed data (up from 3GB) and unlimited calling to more than 100 countries.

  • The $30 Unlimited plan includes 20GB of high-speed data (up from 5GB) and unlimited calling to more than 125 countries (up from 100 countries).

  • The $40 Unlimited plan includes 30GB of high-speed data (up from 15GB) and unlimited calling to more than 125 countries (up from 100 countries).

  • The $50 Unlimited World plan, formerly named the Truly Unlimited plan, includes unlimited high-speed data and unlimited calling to more than 200 countries (a doubling of the number of countries from before).

  • The $60 Unlimited World Plus plan, formerly the Truly Unlimited Plus plan, includes unlimited access to Verizon’s fastest network, 5G Ultra Wideband (where available) and unlimited high-speed data. It also doubles the number of countries with unlimited calling to over 200, and offers international roaming in more than 140 countries.

Total Wireless Plans

The top two Total Wireless plans, Total 5G Unlimited ($50 a month) and Total 5G Unlimited Plus ($60 a month), feature unlimited high-speed data, including 5G Ultra Wideband speeds.

As of Aug. 28, roaming coverage for those plans doubles to more than 30 countries and international calling to 180 countries.

The Total Base 5G Unlimited plan stays the same with unlimited high-speed data (but not 5G Ultra Wideband), international calling to more than 85 countries and texting to over 200 destinations and roaming in Canada and Mexico.

Continue Reading

Trending

Copyright © Verum World Media