Technologies

Bing AI Bungles Search Results at Times, Just Like Google

Expect search engines to tread carefully as tests reveal problems with AI language skills.

A close reading of Bing’s AI-boosted search results reveals that the website can make the same kinds of errors that are apparent in the ChatGPT technology foundation it uses and in Google’s competing Bard.

A new version of Bing that’s in limited testing employs large language model, or LLM, technology from OpenAI, the research lab Microsoft invested in and that grabbed the tech spotlight with its ChatGPT AI chatbot. ChatGPT and related technology, trained on enormous swaths of the internet, can produce remarkable results, but they don’t truly know facts and they can make errors.

Bing tries to avoid such errors by «grounding» results with Microsoft’s search technology, assessing the veracity and authority of source documents and offering links to sources, so people can evaluate the results on their own better. But AI and search engine researcher Dmitri Brereton and others have spotted errors in Bing’s demo, including flawed financial data from a Gap quarterly earnings report.

It’s a buzzkill moment for AI. The technology truly can produce remarkable and useful results, but the trouble is assessing when it isn’t doing that. Expect more caution as the search engine industry tries to find the right formula.

Google felt the pain after a demo last week when its Bard tool, which isn’t yet publicly available, produced erroneous information about the James Webb Space Telescope.

Microsoft offered a similar response Tuesday to what Google said about its gaffe: «We recognize that there is still work to be done and are expecting that the system may make mistakes during this preview period, which is why the feedback is critical so we can learn and help the models get better.»

One basic problem is that large language models, even if trained using text like academic papers and Wikipedia entries that have passed some degree of scrutiny, don’t necessarily assemble factual responses from that raw material.

As internet pioneer and Google researcher Vint Cerf said Monday, AI is «like a salad shooter,» scattering facts all over the kitchen but not truly knowing what it’s producing. «We are a long way away from the self-awareness we want,» he said in a talk at the TechSurge Summit.

Summarizing documents might appear to be within AI’s advanced language processing abilities, but constructing human-readable sentences without drawing in inappropriate information can be difficult. For example, in this request to summarize a spec sheet for Canon’s new R8 mirrorless camera, Bing technology rattled off many features that actually are found in Canon’s earlier R5.

Another high-profile computer scientist, former Stanford University Professor John Hennessy, was more bullish than Cerf, praising AI for its language skills in understanding search queries and generating results that often are correct. But he also cautioned, «It’s always confident that it has the right answer even when it doesn’t.»

OpenAI itself offers the same warning. «ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers. Fixing this issue is challenging,» the AI research lab said when it launched ChatGPT in November.

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

Technologies

Meta Allegedly Profited by $16B From Scam Ads. US Senators Demand FTC, SEC Probe

The lawmakers say Meta’s platforms may be implicated in «about a third of all US scams» and linked to more than $50 billion in consumer losses last year.

US Senators Josh Hawley and Richard Blumenthal are calling for an investigation into Meta over its alleged role in profiting from scam-laden advertisements on Facebook and Instagram. The demand follows a Reuters investigation reportedly based on internal Meta documents that estimated that nearly 10% of Meta’s 2024 revenue — about $16 billion — came from alleged «illicit advertising.»

In a letter to the Federal Trade Commission and the Securities and Exchange Commission, the lawmakers urged regulators to «immediately open investigations and, if the reporting is accurate, pursue vigorous enforcement action … to force Meta to disgorge profits, pay penalties and agree to cease running such advertisements.»

One document reportedly alleged that Meta earned $3.5 billion in just six months from what it classified as «higher-risk» scam ads.

The same internal records reportedly suggested that many ads allegedly violating fraud rules were permitted to run because they «did not apply to many ads… [that staff] believed ‘violated the spirit’ of its rules against scam advertising.»

Meta denies all of these allegations.

Don’t miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source.

Why this matters for you

The scale of this alleged fraud raises serious concerns about Meta’s business model. Many question whether the company is doing enough to police its ad ecosystem, given that a major revenue stream appears to be tied to deceptive or outright fraudulent campaigns.

The senators allege that Meta’s lax enforcement — combined with the continued presence of gambling ads, payment scams, political deepfakes and other dangerous content in its public Ad Library — underscores significant risks.

In their letter, Hawley and Blumenthal highlighted that reducing reports of scam ads by 58% over 18 months — as Meta says — may not tell the whole story. They pointed to broader trends that, according to their own reading of the documents, Meta’s platforms may be allegedly implicated in «about a third of all US scams» and linked to more than $50 billion in consumer losses last year.

What Meta is saying

Meta responded defensively to this call for investigation.

Meta spokesman Andy Stone criticized the senators’ allegations as «exaggerated and wrong,» insisting that the company «aggressively fights fraud and scams because people on our platforms don’t want this content, legitimate advertisers don’t want it, and we don’t want it either.»

Ongoing misinformation on Meta platforms

Among the more provocative allegations is that some of the scam ads impersonate government figures or political leaders. The senators point to specific examples, including a bogus advertisement that falsely claimed President Donald Trump was offering $1,000 to food assistance recipients.

They also raise concerns that foreign cybercrime groups based in countries such as China, Sri Lanka, Vietnam and the Philippines could be behind many of the scam campaigns.

In general, the senators’ push for FTC and SEC action seeks accountability for a social media giant whose ad system may be fueling fraud at an unprecedented scale, though Meta publicly underscores its commitment to user safety. With so much of Meta’s business potentially tied to high-risk ads, the outcome of any investigation could reshape not only its practices but also broader regulatory expectations for major tech companies going forward.

Technologies

Today’s NYT Connections Hints, Answers and Help for Nov. 25, #898

Here are some hints and the answers for the NYT Connections puzzle for Nov. 25, #898

Looking for the most recent Connections answers? Click here for today’s Connections hints, as well as our daily answers and hints for The New York Times Mini Crossword, Wordle, Connections: Sports Edition and Strands puzzles.

Today’s Connections: Sports Edition is both fun and tricky. Think colors, and see if you can find items that share the same color. The purple category wants you to look within four words and find a connecting word hidden in them. If you’re struggling with today’s puzzle but still want to solve it, read on for hints and the answers.

The Times now has a Connections Bot, like the one for Wordle. Go there after you play to receive a numeric score and to have the program analyze your answers. Players who are registered with the Times Games section can now nerd out by following their progress, including the number of puzzles completed, win rate, number of times they nabbed a perfect score and their win streak.

Read more: Hints, Tips and Strategies to Help You Win at NYT Connections Every Time

Hints for today’s Connections groups

Here are four hints for the groupings in today’s Connections puzzle, ranked from the easiest yellow group to the tough (and sometimes bizarre) purple group.

Yellow group hint: Not big.

Green group hint: These things are the same color.

Blue group hint: Itchy!

Purple group hint: They all share words that are positive.

Answers for today’s Connections groups

Yellow group: Small-time.

Green group: Things that are pink.

Blue group: Things you can scratch.

Purple group: Starting with optimistic words.

Read more: Wordle Cheat Sheet: Here Are the Most Popular Letters Used in English Words

What are today’s Connections answers?

The yellow words in today’s Connections

The theme is small-time. The four answers are Mickey Mouse, rinky-dink, trivial and two-bit.

The green words in today’s Connections

The theme is things that are pink. The four answers are Barbie Dreamhouse, calamine lotion, cherry blossom and flamingo.

The blue words in today’s Connections

The theme is things you can scratch. The four answers are bug bite, lottery ticket, vinyl record and your head.

The purple words in today’s Connections

The theme is starting with optimistic words. The four answers are glad-hand, Happy Meal, merry-go-round and sunny-side up.

Technologies

I’m Buying These Headphones for My Gamer Thanks to This Deep Black Friday Discount

SteelSeries headphones are among the top gaming headsets on the market and right now you can score them for as much as half off ahead of Black Friday.

Ready or not, Black Friday is just about here. And Black Friday deals are already popping up on the regular.Here at CNET, we offer our services as dedicated deal finders, scouring the market to curate the best offers for you. We recently spotted some of our favorite gaming headsets on sale at Amazon, some of them for more than half off. We don’t know if these prices will last all the way to Black Friday itself, so if you see something you like, we suggest grabbing it sooner rather than later.

Right now, several SteelSeries’ most popular gaming headsets are seeing nice discounts. I’ve highlighted some of the best offers below, including the ones in my own cart right now, but be sure to check out the Amazon page to see all the options.

First up, the headset in my cart right now as as a holiday gift for my discerning teenage son is the SteelSeries Arctis Nova 3x wireless gaming headset, currently down to just $90, a $20 savings. Arctis Nova 3x comes in blue, black or white and is compatible with Xbox, PC, PS5, Switch and mobile and has a whopping 40-hour battery life. They are lightweight and have a stretchy headband for added comfort.

If you want to go pro, check out the SteelSeries Arctis Nova Pro wireless headset, down to $290 from $380. That’s nearly $100 off a multi-platform compatible headset good for Xbox, Switch, PS5, PS4, PC and mobile. This headset was our sibling site, PCMag’s top pick for best overall gaming headphones. It boasts active noise cancellation, 360-degree spatial audio and a ClearCast Gen 2 mic. Only the black variant is currently on sale at this big of a discount.

Essentially a more affordable version of the above, the Arctis Nova Pro wired headset for Xbox is down to $150, which is $100 off the regular price of $250. This one also got PCMag’s seal of approval for best wired gaming headphones.

Buyers should note that a few of the headsets on sale are renewed, so be sure to double check before buying. We’ve also rounded up the best Black Friday headphone deals if you’re in the market for something for an audiophile near you.

CHEAP GAMING LAPTOP DEALS OF THE WEEK

Alienware 16 Aurora AC16250 Intel 7 RTX 4050 512GB SSD 16GB RAM 16-inch 120Hz Laptop:

$900 (save $200)
Lenovo LOQ Intel i5 13th Gen RTX 5050 1TB SSD 16GB RAM 15.6-inch Laptop:

$800 (save $245)
HP Omen Slim Intel Ultra 5 RTX 5050 512GB SSD 16GB RAM 16-inch Laptop:

$950 (save $500)
Alienware 16 Aurora AC16250 Intel 7 RTX 5050 512GB SSD 16GB RAM 16-inch 120Hz Laptop:

$1,000 (save $200)
HP Victus Ryzen 7 RTX 4050 1TB SSD 16GB RAM 15.6-inch Laptop:

$900 (save $300)

Deals are selected by the CNET Group commerce team, and may be unrelated to this article.

Why this deal matters

Gaming headphones come in a wide range of styles and price points, but the best ones factor in three key components: sound quality, battery life and comfort — those all-night gaming sessions pretty much demand it. Over the years, we’ve tested a ton of brands, and SteelSeries frequently comes out on top, especially for PC and Xbox users. Being able to pick up a pair at a discount makes it a great opportunity to gift or grab a new headset that delivers on comfort, sound and quality.

Join Our Daily Deals Text Group!

Get hand-picked deals from CNET shopping experts straight to your phone.

By signing up, you confirm you are 16+ and agree to receive recurring marketing messages at the phone number provided. Consent is not a condition of purchase. Reply STOP to unsubscribe. Msg & data rates may apply. View our Privacy Policy and Terms of Use.