Author: Admin

There are plenty of different ways to play Wordle, The New York Times’ engaging online word-guessing game. Personally, I like to stick with starter words that I know feature some of the most frequently used letters in the English language. (TRAIN and CLOSE are among my favorites.)

But exactly which letters are the most commonly used? I reached out to the big brains at the Oxford English Dictionary — you know, the whopping, multi-volume dictionary that no one actually owns, but that your school librarian adored. An OED representative sent me a list in which staff analyzed all of the main entries in the concise version of the dictionary.

“It’s worth noting that the piece is pretty old,” the OED representative said apologetically. “It mentions that the stats are from Concise Oxford English Dictionary in 2004, although current colleagues aren’t sure when the piece itself was written, and I’m afraid we don’t have any newer analysis of this sort.”

The list may be 20 years old, but I’m going with it. The alphabet isn’t like ice cream flavors or musical acts — no new, trendy letters have been introduced in that time. So whether you’re just getting into Wordle, or you want to shake up your starter words and make sure they’re still hitting the most common letters, this list is for you.

S begins the most English words

The most common letter is no surprise to most people: It’s E, and it’s immediately followed by another vowel, A. And those letters are much, much more popular than the final two letters in the OED’s list, which are J and Q. (Anyone else surprised that J is used less than X and Z?)

“The letter E is over 56 times more common than Q in forming individual English words,” the OED analysis found.

It notes that “there are more English words beginning with the letter S than with any other letter.”

So Wordle players, bookmark this article, and you might find yourself doing a little better in upcoming games. Here are the letters of the alphabet ranked according to how frequently they showed up in the 2004 Concise Oxford English Dictionary.

The most common letters used in English

1. E

2. A

3. R

4. I

5. O

6. T

7. N

8. S

9. L

10. C

11. U

12. D

13. P

14. M

15. H

16. G

17. B

18. F

19. Y

20. W

21. K

22. V

23. X

24. Z

25. J

26. Q

Today’s Wordle answer could trip you up. It’s not an obscure word, but it’s not one I thought qualified for Wordle. If you’re stuck, read on.

Every day, we’ll post hints and then the answer for the current day’s Wordle, just in case you need it.

Today’s Wordle hints for April 25

Warning: If you keep reading, you’ll see the Wordle answer for Thursday, April 25, puzzle No. 1041. That could be a devastating spoiler for some players. But if you just need the answer — maybe you’re on your last guess and just don’t want to see an 800-game streak go poof — keep reading.

Wordle hint No. 1: Vowels

Today’s Wordle answer has two vowels.

Wordle hint No. 2: Consonants

That leaves three consonants, and oddly, they’re all in the middle of the word, in spaces two, three and four.

Wordle hint No. 3: Starter letter

Today’s Wordle answer begins with the letter I.

Wordle hint No. 4: Abbreviated

Today’s Wordle answer is short for a longer word.

Wordle hint No. 5: Meaning

Today’s Wordle answer refers to the beginning of something, such as a book or piece or music.

TODAY’S WORDLE ANSWER FOR APRIL 25

Here comes the spoiler: Today’s Wordle answer is INTRO. I admit it, I was surprised the Wordle editors used this word, because I think of it as an abbreviation or slang for “introduction” and for some reason expected that would make it off-limits for Wordle. But after all, there are only so many five-letter words out there.

Yesterday’s Wordle answer

Yesterday’s answer, #1040, was OVERT.

Past Wordle answers

April 20, No. 1036: LUCID

April 21, No. 1037: JOLLY

April 22, No. 1038: LASER

April 23: No. 1039: ROVER

Everyday Wordle tips

I’ve written a lot about Wordle — from covering its 1,000th word to my list of the best starter words to a helpful two-step strategy to news about controversial word changes. I’ve even rounded up what I learned playing the hit online word puzzle for a full year. So if you’re rethinking your need for the actual answer, you might try tips from one of those stories.

Still need a starter word? One person told me they just look around and choose a five-letter object that they’ve spotted to use as their starter word — such as COUCH or CHAIR. I tend to stick to starter words that have the most popular letters used in English words. I like TRAIN as a starter, though I have a friend who uses TRAIL. I’ve read that people use the financial term ROATE, but I like to use words I actually know.

What is Wordle?

If you read this far, you know how to play. You have six chances to guess a five-letter word, and the game gives you feedback as to whether the letters you’ve guessed are in the puzzle, and if they’re in the same spot where you guessed they are. The New York Times bought the game from creator Josh Wardle for seven figures in 2022. Wardle famously created the game for his partner, and let her narrow down the 12,000 five-letter words in the English language to just 2,500, creating the database of answers. It was convenient, too, that he has a name that plays off “word.”

A Times spokesperson told me it lists the very first Wordle as appearing on June 19, 2021, and the paper celebrated the game’s 1000th word on March 15. Spoiler: That 1000th word was ERUPT. As in, “Dad will erupt if the Wordle is so tough he loses his streak.”

There are other fun games in the Times Games stable. My latest addiction is Connections, which I think is trickier than Wordle. This is the game where you look at a grid of 16 words and try to put them into four groups of related words. Sometimes the relationships between the words are pretty out there — like the time when it was four words that all began with rock bands, such as “Rushmore” and “journeyman.” (Connections got a little sassy on April Fools’ Day with an all-emoji puzzle. Some gamers did NOT find that funny.)

Spelling Bee is a popular Times game too. And there’s a new game that’s still in beta, Strands, which I’m trying to master.

When you ask an AI chatbot like ChatGPT, Claude, Copilot or Gemini to do something, it may seem like you’re interacting with a person.

But you’re not. These chatbots don’t actually understand the meaning of words the way we do. Instead, they’re the interface we use to interact with large language models, or LLMs. This underlying technology is trained to recognize how words are used and which words frequently appear together, so it can predict future words, sentences or paragraphs.

Generative AI tools are constantly refining their understanding of words to make better predictions. Some, including Google’s Lumiere and OpenAI’s Sora, are even learning to generate images, video and audio.

It’s all part of a constant flux of one-upmanship kicked off by ChatGPT’s introduction in late 2022, followed by the arrival of Microsoft’s AI-enhanced Bing search and Google’s Bard (now Gemini). Over the ensuing months, Microsoft introduced Copilot, Meta updated Llama, OpenAI released Dall-E 3 and GPT-4 Turbo, Google announced Gemini Ultra 1.0 and teased Gemini 1.5 Pro, while Anthropic debuted Claude 3. Google and Adobe have released peeks at tools that can generate virtual games and music to show consumers where the technology is headed.

Cutting-edge technology like this has arguably never been so accessible. And the companies developing it are eager to lure you into their ecosystems and to stake their claims in a market projected to be worth $1.3 trillion by 2032.

If you’re wondering what LLMs have to do with AI, this explainer is for you. (And be sure to check out our new AI Atlas guide for hands-on product reviews, as well as news, tips, video and more.)

What is a language model?

You can think of a language model as a soothsayer for words.

“A language model is something that tries to predict what language looks like that humans produce,” said Mark Riedl, professor in the Georgia Tech School of Interactive Computing and associate director of the Georgia Tech Machine Learning Center. “What makes something a language model is whether it can predict future words given previous words.”

This is the basis of autocomplete functionality when you’re texting, as well as AI chatbots.

What is a large language model?

A large language model is, by definition, a big language model.

How big?

These models are measured in what is known as “parameters.”

What’s a parameter?

Well, LLMs use neural networks, which are machine learning models that take an input and perform mathematical calculations to produce an output. The number of variables in these computations are parameters. A large language model can have 1 billion parameters or more.

“We know that they’re large when they produce a full paragraph of coherent fluid text,” Riedl said.

Is there such a thing as a small language model?

Yes. Tech companies like Microsoft are rolling out smaller models, designed specifically for phones and PCs, that don’t require the same computing resources as an LLM but nevertheless help users tap into the power of generative AI.

How do large language models learn?

LLMs learn via a process called deep learning.

“It’s a lot like when you teach a child — you show a lot of examples,” said Jason Alan Snyder, global CTO of ad agency Momentum Worldwide.

In other words, you feed the LLM a library of content (what’s known as training data) such as books, articles, code and social media posts to help it understand how words are used in different contexts — and even the more subtle nuances of language.

During this process, the model digests far more than a person could ever read in their lifetime — something on the order of trillions of tokens.

Tokens help AI models break down and process text. You can think of an AI model as a reader who needs help. The model breaks down a sentence into smaller pieces, or tokens — which are equivalent to four characters in English, or about three-quarters of a word — so they can understand each piece and then the overall meaning.

From there, the LLM can analyze how words connect and determine which words often appear together.

“It’s like building this giant map of word relationships,” Snyder said. “And then it starts to be able to do this really fun, cool thing, and it predicts what the next word is … and it compares the prediction to the actual word in the data and adjusts the internal map based on its accuracy.”

This prediction and adjustment happens billions of times, so the LLM is constantly refining its understanding of language and getting better at identifying patterns and predicting future words. It can even learn concepts and facts from the data to answer questions, generate creative text formats and translate languages. But they don’t understand the meaning of words like we do — just the statistical relationships.

LLMs also learn to improve their responses through reinforcement learning from human feedback.

“You get a judgment or a preference from humans on which response was better given the input that it was given,” said Maarten Sap, assistant professor at the Language Technologies Institute at Carnegie Mellon. “And then you can teach the model to improve its responses.”

What do large language models do?

Given a series of input words, a LLM can predict the next word.

For example, consider the phrase, “I went sailing on the deep blue…”

Most people would probably guess “sea” because sailing, deep and blue are all words we associate with the sea. In other words, each word sets up context for what should come next.

“These large language models, because they have a lot of parameters, they can store a lot of patterns,” Riedl said. “They are very good at being able to pick out these clues and make really, really good guesses at what comes next.”

What do large language models do really well?

LLMs are very good at figuring out the connection between words and producing text that sounds natural.

“They take an input, which can often be a set of instructions, like, ‘Do this for me’ or ‘Tell me about this’ or ‘Summarize this’ and are able to extract those patterns out of the input and produce a long string of fluid response,” Riedl said.

Where do large language models struggle?

But they have several weaknesses.

First, they’re not good at telling the truth. In fact, they sometimes just make stuff up that sounds true, like when ChatGPT cited six fake court cases in a legal brief or when Bard mistakenly credited the James Webb Space Telescope with taking the first pictures of a planet outside of our own solar system. Those are known as hallucinations.

“They are extremely unreliable in the sense that they confabulate and make up things a lot,” Sap said. “They’re not trained or designed by any means to spit out anything truthful.”

They also struggle with queries that are fundamentally different from anything they’ve encountered before. That’s because they’re focused on finding and responding to patterns.

A good example is a math problem with a unique set of numbers.

“It may not be able to do that calculation correctly because it’s not really solving math,” Riedl said. “It is trying to relate your math question to previous examples of math questions that it has seen before.”

And while they excel at predicting words, they’re not good at predicting the future, which includes planning and decision making.

“The idea of doing planning in the way that humans do it with … thinking about the different contingencies and alternatives and making choices, this seems to be a really hard roadblock for our current large language models right now,” Riedl said.

Finally, they struggle with current events because their training data typically only goes up to a certain point and anything that happens after that isn’t part of their knowledge base. And because they don’t have the capacity to distinguish between what is factually true and what is likely, they can confidently provide incorrect information about current events.

They also don’t interact with the world the way we do.

“This makes it difficult for them to grasp the nuances and complexities of current events that often require an understanding of context, social dynamics and real-world consequences,” Snyder said.

How will large language models evolve?

We’re already starting to see generative AI companies like OpenAI and Adobe debut multimodal models, which are trained not just on text but on images, video and audio.

We’ll also likely see improvements in LLMs’ abilities to not just translate languages from English but to understand and converse in additional languages as well.

We may also see retrieval capabilities evolve beyond what the models have been trained on. That could include leveraging search engines like Google so the models can conduct web searches and then feed those results into the LLM.

If LLMs were connected to search engines, they could process real-time information far beyond their training data. This means they could better understand queries and provide more accurate, up-to-date responses.

“This helps our linkage models stay current and up to date because they can actually look at new information on the internet and bring that in,” Riedl said.

There are a few catches. Web search could make hallucinations worse without adequate fact-checking mechanisms in place. And LLMs would need to learn how to assess the reliability of web sources before citing them. Plus, it would require a lot of (expensive) computing power to process web search results on demand.

AI-powered Bing, which Microsoft announced in February 2023, is a similar concept. However, instead of tapping into search engines to enhance its responses, Bing is using AI to make its own search engine better. That’s in part by better understanding the true meaning behind consumer queries and better ranking the results for said queries.

Editors’ note: CNET is using an AI engine to help create some stories. For more, see this post.

Whether you’re around the pool or the campfire, a great Bluetooth speaker is an absolute must-have. Listening to music, podcasts, audiobooks or anything else doesn’t have to mean using a pair of earbuds or headphones, and there are tons of portable speakers on the market today. But few look quite so cool as the Soundcore Motion X500, a portable speaker that comes in a fetching purplish pink and has a retrolike look that somehow still fits in 2024’s world. And even better, you can pick one up today for just $100 when you clip the on-page coupon. Note that the blue and black versions of this speaker aren’t included in this offer, so you do have to go pink — thankfully, that’s the best one!

This speaker does more than just look cool, though. It’s certified for wireless hi-res sound, with what the company says is a three times more detailed listening experience, and it supports a Soundcore algorithm that’s designed to transform stereo sound into a three-channel spatial audio experience. The result is sound that appears to come from more than just the speaker itself, great for your favorite tunes.

Other features of note include a 12-hour-long battery life and an IPX7 water resistance rating, which means you can splash this thing by the pool and also never worry about it getting caught in a rainstorm. But remember, you do need to clip that code to take advantage of this particular Bluetooth speaker deal. We don’t know how long the offer will stick around, so bear that in mind as well.

Prefer to keep your audio a little more personal? Our list of the best earbud and headphone deals will be right up your street.

A great tablet can help you get work done, binge-watch your favorite content and stay organized. If you’ve been looking to acquire one for yourself, then this tablet might be the right fit. With its ample screen size, reliable performance and beautiful graphics, you’ll love this deal. As of today, you can snag the Lenovo Tab Extreme for only $950 at Best Buy. You’ll save $150 off its regular price of $1,100.

The Lenovo Tab Extreme is a 14.5-inch powerhouse with 12GB RAM for speedy performance and 256GB memory so you can store your important documents, as well as your favorite software and apps. But there’s more: this deal also includes a keyboard and precision pen so you can create without having to worry about spending on accessories that will make it easier to get work done.

Need to watch your favorite shows or look at presentations? The OLED display sports a 3,000×1,876 screen resolution that should make colors pop and enhance contrast, which will come in handy if you’re video editing. Its screen size and the included accessories really make this tablet a great laptop substitute or addition to your home office arsenal. The Tab Extreme includes a 13 MP front-facing camera and a 13 MP + 5 MP rear camera that lets you make videos and look your best during work or personal video calls.

The sound should also be solid thanks to the octa-JBL acoustic system. Plus, you’ll be able to download everything the Google Play store has to offer and easily sync this tablet to your other Android devices. At only $950, this tablet, pen and keyboard deal is a steal.

Looking for a tablet but not sure if this is the one for you? Check out other CNET-reviewed recommendations.

The iPhone 17 Plus will have a smaller screen than the 6.7-inch display on the current-generation iPhone 15 Plus, according to a rumor posted on X last week by longtime supply-chain analyst Ross Young.

The leaked information on the iPhone 17, reported earlier by MacRumors, comes before Apple’s iPhone 16 has even been announced. The iPhone 17 lineup would likely launch in 2025.

Apple didn’t immediately respond to a request for comment.

Apple changing the screen size of its iPhones wouldn’t be much of a surprise, though it is noteworthy. The company has always seemed comfortable fiddling with screen sizes, particularly as its product lineup expanded from one iPhone each year to multiple models, starting with the iPhone 6 and iPhone 6 Plus in 2014.

It’s unclear why Apple would opt for a smaller-size screen for its iPhone 17 Plus, but MacRumors suggests the change could help further differentiate the Plus model from its more expensive Pro Max cousin. Though Apple is expected to increase the screen size of its upcoming iPhone 16 Pro and iPhone 16 Pro Max this fall, changing the screen size on the iPhone 17 Plus would likely help each iPhone stand apart even more.

It’s also possible Apple has learned through customer feedback that customers who look at the Plus-size iPhones want a device that’s bigger than the entry-level iPhone but not as big as the Pro Max.

Read more: iPhone 16: Major Rumors About Apple’s Next iPhone

As mentioned, Apple is expected this year to adjust screen sizes of the iPhone 16 Pro and iPhone 16 Pro Max, which will reportedly have slimmer bezel borders and larger displays. So it’s only natural to believe Apple would then focus its screen-fiddling energy on the entry-level iPhone 17 and iPhone 17 Plus next year.

You might be tempted to think Apple has this screen-size thing all worked out. Currently, Apple’s lineup seems rather fixed at a $799 entry-level iPhone, followed by a larger entry-level iPhone Plus for $899, and then a $999 Pro level, with larger Pro Max versions starting at $1,099. But Apple only stumbled on that formula after attempting to sell a 5.4-inch iPhone Mini as far back as 2021.

The iPhone Mini earned positive reviews from critics, including CNET’s Patrick Holland, who wrote in 2020 that the iPhone 12 Mini was the “small phone we’ve been asking for,” adding that it was “a one-handed phone user’s dream.” Alas, Apple switched to larger entry-level phones with the iPhone 14 line in 2022.

I Took 600+ Photos With the iPhone 15 Pro and Pro Max. Look at My Favorites

+31 More

See all photos

President Joe Biden on Wednesday signed into law legislation that could lead to an eventual ban of TikTok, though the popular social media platform and others have vowed to fight the measure in court.

The US Senate voted 79-18 to approve the bill late Tuesday as part of a $95 billion aid package for Ukraine, Israel and other US allies. It gives TikTok’s Chinese owner, ByteDance, up to a year to sell TikTok to a buyer deemed fit by US officials.

In a Wednesday TikTok post, Shou Chew, the company’s CEO, called the signing of the law “disappointing” and said TikTok will fight it in court.

“Rest assured, we aren’t going anywhere,” Chew said. “We are confident and will keep fighting for your rights in the courts. The facts and Constitution are on our side and we expect to prevail again.”

The House originally passed legislation in March that would’ve led to a ban unless TikTok was sold off within six months, but the measure failed to gain traction in the Senate. The current version would extend that time period to nine months, with a potential for an additional three-month extension.

Lawmakers in both political parties have long voiced concerns that the popular video app, which has more than 150 million American users, could be a threat to national security and could be used by the Chinese government to spy on Americans or spread disinformation to further China’s agenda.

TikTok continues to deny those accusations. Ahead of the March vote on the previous legislation, TikTok rallied its US users, calling on them to urge their representatives on Capitol Hill to vote it down.

Experts say the unprecedented move will undoubtedly prompt legal challenges from free speech advocates, the tech industry and others, especially in the absence of any direct evidence showing Chinese government ties or surveillance.

So what’s next for lawmakers and TikTok? Here’s what you need to know.

What would the bill do?

The bill is aimed at forcing ByteDance to sell TikTok to a buyer that American officials are OK with, as well as guaranteeing that ByteDance no longer has access to US user data or control over the TikTok algorithm that decides what videos American users see.

If TikTok doesn’t comply within nine months of the bill taking effect, the government could require the removal of its app from US app stores. Under the new legislation, that time period could potentially be extended to a full year.

Read more: TikTok Loves to Give Financial Advice. But Don’t Believe Everything You Hear

What’s next?

While Biden pledged to sign the original bill if it made it through Congress, former President Donald Trump, the presumptive Republican nominee running against Biden in November, has said he doesn’t support a ban.

After originally calling for a ban during his presidency, Trump said in March on CNBC’s Squawk Box that though he still thinks the app is a danger to national security, he no longer thinks it should be banned, saying that “there are a lot of young kids on TikTok who will go crazy without it.”

Trump went on to say that banning TikTok would only boost the power of Facebook, which he referred to as an “enemy of the people.”

Read more: Everything You Need to Make Better TikTok Videos

Who else opposes the bill?

Free speech and digital rights groups, as well as some security experts, have long opposed the idea of a ban, saying that singling out TikTok doesn’t do anything to solve the broader problems with social media as a whole.

Instead, they argue that lawmakers would be better off passing comprehensive digital privacy laws that would protect the personal information of Americans by stopping all social media companies from collecting it and selling it to data brokers.

In a blog post, the Electronic Frontier Foundation said in March that until that happens, even if TikTok is banned or sold, there’s nothing stopping the Chinese government, or others, from just buying up that same data.

“Ultimately, foreign adversaries will still be able to obtain our data from social media companies unless those companies are forbidden from collecting, retaining and selling it, full stop,” the EFF said.

Digital privacy rights group Fight for the Future agreed. In a Tuesday statement its director, Evan Greer, called the bill “one of the stupidest and most authoritarian pieces of tech legislation we’ve seen in years.”

“Not only is this bill laughably unconstitutional and a blatant assault on free expression and human rights, it’s also a perfect way to derail momentum toward more meaningful policies like privacy and antitrust legislation that would actually address the harms of big tech and surveillance capitalism,” Greer said.

Greer added that banning TikTok in the absence of data privacy regulation will only solidify the monopolies of Meta and Google, without doing anything to protect Americans from data harvesting or government propaganda.

Samsung has created TVs that can hold their own against the glare of bright rooms, making it possible to enjoy your favorite shows without having to deprive yourself of natural light. If you’ve been looking for a reliable OLED TV, today is your lucky day. Samsung is offering discounts on its trusty OLED S95D series TVs right now, meaning you can snag a 55-inch OLED 95D for only $2,300, down from its regular price of $2,600.

Samsung has equipped this series of TVs with OLED technology that lets you see pure blacks and bright whites, bringing your favorite films, shows and video games to life. Anti-glare technology lets you enjoy your TV binges during the day without pesky visual disturbances that can prevent you from seeing every last detail. Additionally, these TVs also feature OLED’s HDR pro, which improves brightness so you can have an immersive experience.

Gamers will also enjoy the S95D’s Motion Xcelerator 144Hz, a feature that reduces lag and blurriness. But Samsung also designed this TV with Dolby Atmos and Object Tracking Sound Plus. These features provide a multichannel sonic experience even as objects on the screen move. S95D TVs include the NQ4 AI Gen2 Processor, an AI processor that powers every feature included in this TV.

Here’s a list of every Samsung OLED S95D on sale today in addition to the 55-inch model:

• Samsung OLED S95D, 65-inch: $3,000 (save $400)
• Samsung OLED S95D, 77-inch: $4,300 (save $300)

Amazon and Best Buy are matching these specials today as well. Best Buy is also including free installation for TVs over 56 inches, which amounts to a discount of $250 and added peace of mind.

Would you be more likely to buy a phone if you knew that you’d be able to repair it at home if you smashed the screen? Finland-based phone-maker HMD is hoping you will, as it unveils its first own-brand phone series designed around the concept of repairability.

HMD is best known for making Nokia phones — both the smart and dumb varieties — but now it’s branching out by making phones with its own name stamped on the back. On Wednesday it unveiled its first three own-brand devices, the HMD Pulse, Pulse Pro and Pulse Plus. The Pulse, priced at 100 (around $125 or AU$190 converted), and the Pulse Plus, priced at 130 ($160, AU$250) are available to purchase immediately in some markets, whereas the Pulse Pro is currently listed as “coming soon.”

The most exciting of the three devices is the HMD Pulse Pro, which has 50-megapixel cameras on both the back and front, meaning that your selfies will be the same quality as the photos you take with the main camera. The three phones also share many features and attributes, including a Unisoc T606 processor, Android 14, and a focus on affordability and repairability. With all three devices, you can replace a damaged display, bent charging port or a depleted battery in just a few minutes using simple tools.

The ability to repair your own device at home is an increasingly desirable feature in new tech hitting the market. Extending the life of products means you need to shell out less frequently to replace tech that otherwise would be discarded, which in turn reduces e-waste and causes less harm to the environment. It’s a win-win for consumers and for the planet.

Read more: Fairphone 5 Review: The Phone That Wants to Save the World

Some tech companies, including Apple and Samsung, run their own repair programs, but there’s a growing movement to allow people to repair their own devices. Online community, advocacy group and parts retailer iFixit has been a staunch advocate of the right-to-repair movement, and over the past few years has started partnering with tech companies, including HMD, to allow people to buy affordable repair kits specific to their tech.

“HMD’s ongoing commitment to designing phones with repairability in mind is an encouraging shift in our throwaway culture, proving that HMD has the innovative culture needed to drive advances in repairability,” said iFixit CEO Kyle Wiens in a statement. “We’re keen to collaborate as they work towards making repairable phones the norm, not the exception.”

In 2023, one in every four phones HMD sold was repairable. This year, the company is aiming to make that three in every four. For the HMD Pulse, Pulse Pro and Pulse Plus, owners will be able to buy a self-repair kit and spare parts from iFixit starting at 18 ($22 or AU$35). It’s a continuation of the partnership between the two companies that’s also seen repair kits available for HMD’s Nokia smartphones.

The big difference between HMD’s own-brand phones and its Nokia devices seems to be in the design. The Pulse phones are more angular and feature a more pronounced camera module. They also come in a range of fun but muted colors, including “twilight purple.”

“There is an audience who would buy whatever phone just because it’s repairable,” HMD’s CMO Lars Silberbauer told me in February this year, when I spoke to him at MWC in Barcelona and the company first announced it would be making its own phones. “But if you want to reach the masses, then you need to have something that’s also desirable.”

The Pulse series is HMD’s first attempt to create a desirable and repairable phone that aims to attract Gen Z as it begins its journey as a fledgling phone brand. It’s a crowded market, but being a step ahead where it comes to repairability might just give HMD a competitive edge.