Stats, iPhones, and AI devices are still a threat


In this week’s “Sunday Reboot,” scientists somehow say iPhones lowered birthrates, and Musk’s AI prototype story shows that specialist hardware is coming, even if we don’t necessarily want it.

Sunday Reboot is a weekly column covering some of the lighter stories within the Apple reality distortion field from the past seven days. All to get the next week underway with a good first step.

iPhones up, birthrate down, “research” questionable

The iPhone is a mere 19 years old this week, first going on sale on June 29, 2007. It’s old enough to drink in the UK, if not quite in the United States.

Before it flies across the Atlantic to buy a brew from a bar in Basingstoke, some pesky scientists have tried to spoil the party.

Raised on July 1, a working paper from the National Bureau of Economic Research proposes “Is the iPhone Birth Control?” If we follow Betteridge’s law of headlines, the answer is “No.”

The paper proclaims “National-survey evidence on time use and sexual behavior is consistent with the iPhone reducing in-person interactions, increasing pornography use, and reducing sexual frequency” in the United States

Five firstgeneration iPhones standing side by side on a black background, each displaying different screens such as maps, web browser, home screen, music, and phone call, with text Say hello to iPhone

Apple launched the iPhone in 2007. Scientists say it was too hot for the U.S. Birth Rate – image credit: Apple

Hearing that for the first time, you’d nod and agree with it. That song from the puppet musical “Avenue Q” is right, and so is the abstract.

However, AppleInsider William Gallagher looked into the report and realized that there are some catches to this assumption.

First, that it’s not really a phenomena associated with just the iPhone. I recall there were many different other devices on the market at the time that were capable of being used in illicit ways.

However, the writers decided to focus on data of AT&T users between 2005 and 2011. That’s two years before the iPhone and the years when it was exclusive to AT&T.

Except it doesn’t really take into account the fact that AT&T also sold many other early smartphones too. Blackberries, the HTC-based AT&T Tilt, the Nokia N75, and others were all buyable from the carrier, not just iPhones.

Even the report writers acknowledge this, saying they don’t claim the iPhone was the sole cause of the post-2007 decline.” Indeed, they say the estimates imply that the modern smartphone played a “sizable role” in the decline of US births.

But, before the puritan complainers can get a word in about how smartphones are evil, and demand retribution from the House of Jobs, the report goes on to work against the implied statement.

Line graph titled Search interest in porn, showing Google Trends index rising from 2004, peaking around 2013 near 100, then gradually declining and stabilizing between 40 and 60 by 2024

A graph used as proof the iPhone affected the U.S. birth rate. Except it doesn’t… – Image credit: National Bureau of Economic Research

Searches for “porn” shot up in 2014, but have gone down to almost 2009 levels again by 2024. Almost as if society had the novelty of trying to access specialist content wherever they want, then the novelty wore off.

It’s an odd report that tries to prove a point, but works against itself in many ways, and with too many asterisks to really be taken seriously.

In which case, was it really a clickbait way to try and secure some grant funding during some downtime? Or maybe some scientists placed bets that they could get it published.

Maybe it was an excuse to get to look at adult material while working without needing to face an investigation from HR.

Either way, Betteridge’s law is maintained once again.

Don’t stop believing in AI devices

People working in the tech editorial space, such as this very publication, are all too familiar with the notion of prototypes being shown to others. Especially if it’s for a future product in a category where there’s no real dominant option.

Cue a story about one of the Elon Musk companies supposedly sharing a prototype with investors for an AI device. SpaceX’s xAI apparently did that, though Musk has since called the claim “utterly false.”

Your mileage may vary on whether or not to blame the extremely rich man, but he did threaten to go all Futurama Bender on the iPhone in 2022. As someone in charge of tech companies with massive resources capable of doing just that, there was every chance that Musk would do it.

He didn’t, but the opportunity was there.

What the modern-day report provides in detail is minimal. It was shown to investors and was thinner than an iPhone.

While Musk’s insistence that it’s not happening puts a downer on things, it does, however, mean that there’s still an appetite for something to happen in that area.

Three small square wearable camera devices, one white in front and two black behind, each with rounded corners and a horizontal pill-shaped lens area featuring a yellow accent on the left

The Humane AI pin failed. Won’t stop others from doing the same thing – Image Credit: Humane

After the Humane AI pin’s crash and burn and the Rabbit R1’s lack of popularity with consumers, companies should be treading carefully. Those were two high-profile AI device failures, and no-one wants to pay for the third.

Rumors and patents certainly point to Apple continuing down the path of a more AI-focused future. Sure, there’s the questionable pendant, and the ever-present smart glasses, but it’s also looking at non-standalone items too.

We have had rumors insisting that the AirPods will get cameras, providing an environmental view for Apple Intelligence running on an iPhone.

The key bit here is that it’s all feeding back to an iPhone. Something that is obviously a device that has a lot of AI prowess, and could do the same job as Humane’s pin if you push it.

The only difference is that AirPods with cameras, or smart glasses with cameras, would enable that sort of functionality without needing to touch the iPhone at all.

The AirPods don’t have to be “smart” in their own right. They only have to pass along data to the iPhone as the actual “brain” of the operation.

You don’t really need a dedicated device, just a way to interface with what already exists.

So, with the hardware side “solved,” there’s the actual AI bit to cover. Again, Apple Intelligence on the iPhone comes to the rescue, but there is another way.

Humane was painful, particularly since you had to pay a $24 per month subscription for the AI bit. We may not have liked paying a subscription for questionable AI back in 2023, but in 2026, we’re intimately aware of how much people are spending on access to ChatGPT and Claude.

Weirdly, it only took a few years and society is perfectly fine with one of the biggest stumbling blocks that Humane ran into.

It’s not hard to imagine OpenAI releasing hardware that taps into ChatGPT’s paid services. You don’t even have to imagine, as your friendly LLM can tell you about the hardware work with Jony Ive.

Humane flopped so OpenAI and others can walk, run, or be a patronizing yes-man that really wants to explain your assumptions back at you with a more academic style.

Musk may deny the existence of an xAI hardware prototype now, but it would be astounding if one wasn’t being actively developed.

Putting Grok access on a device like that isn’t hard to imagine. For Musk, with an eye staring at his finance teams, it seems like a natural progression of the form, and he gets his own iPhone rival to boot.

Just hope that we don’t get headlines in 20 years time about Musk’s prototype lowering the birth rate again.

Last week’s Sunday Reboot talked about Apple TV and its fear of gameshows. That, and a reminder that it is also bad at making gameshows.



Source link

Leave a Reply

Subscribe to Our Newsletter

Get our latest articles delivered straight to your inbox. No spam, we promise.

Recent Reviews


1,000W, 10-port charger for $45... predictably disappointing.

1,000W, 10-port charger for $45… predictably disappointing. 

Adrian Kingsley-Hughes/ZDNET

Follow ZDNET: Add us as a preferred source on Google.


ZDNET’s key takeaways

  • Things that look “too good to be true” invariable are just that.
  • This example got dangerously hot in a short period of time before dying. 
  • There’s no legitimate charger that comes close to delivering on the 1,000W promise.

Being a tech reviewer for a living means that I get offered some very interesting things. Not interesting as in Bugatti supercars or jewel-encrusted Fabergé eggs, but interesting as in “this thing could easily be a fire hazard — want to take a look?”

Also: The best GaN chargers of 2026: Expert tested

Submissively, I often say yes. And I’m glad I did with the most recent pitch, because it was very interesting indeed.

Meet the “interesting” charger

This time around, the thing of interest was a charger that claimed to deliver an incredible 1,000W through its ten ports — four 140W USB-C ports, four 100W USB-C ports, and two 20W USB-A ports. 

The person who bought this charger told me that they’d plugged it in, used it to charge their phone for “a few minutes,” got worried when it became “a little hot,” and unplugged it.

That's a lot of promise... but (spoilers), they don't deliver!

That’s a lot of promise… but (spoilers), they don’t deliver!

Adrian Kingsley-Hughes/ZDNET

The unit was suspiciously light and plasticky, especially given its built-in power supply. Compare this to Ugreen’s Nexode 500W charger, which weighs a hair under 5 lb.

There was also a slight whiff of melty plastic, which made me think that this had been a bit more than a little hot. 

Also: This $4 router reboot timer is the cheap internet fix I didn’t know I needed – and it works reliably

Color me suspicious, but I had a gut feeling that the only way this charger would be able to push out 1,000W would be if it caught fire. 

Turns out I wasn’t far wrong.

How long would it last? Answer: Minutes

Talk is cheap. It was time to test the charger. 

So I plugged it in, turned it on, and started using it. Within a couple of minutes of starting to use it, I noticed a few things:

  • No matter what I tried, I couldn’t persuade the charger to deliver more than about 60W from any of the ports. 
  • As for peak output, I managed to get close to 250W.
  • The power output was very uneven and noisy, fluctuating wildly. The more ports I used, the worse it got.
  • The unit got very hot to the touch very quickly, even under light loads. 
  • But… before I could get the thermal camera out to check how hot it got, there was a pop and the unmistakable smell of “Magic Smoke.” The charger had been sent to Silicon Heaven within minutes.

Annnnd… POP! This is the moment the charger gave up the ghost.

Adrian Kingsley-Hughes/ZDNET

Diagnosis time

Time to take it apart and have a look inside. For an item that plugged into the mains power, this unit was shockingly easy to take apart. 

A thin sheet of easily removable plastic is a that separates curious hands from live AC power.

A thin sheet of easily removable plastic is a that separates curious hands from live AC power.

Adrian Kingsley-Hughes/ZDNET

And even unplugged and broken, it was capable of delivering zaps! If the case came off while this was plugged into an outlet, it could very easily be deadly.

There’s charge still in some of the capacitors, and these could deliver quite a zap despite the unit being broken and unplugged!

Adrian Kingsley-Hughes/ZDNET

After getting inside, the unit was filled with a grey goo that I’d seen in a previous disappointing charger I’d taken apart. This is a thermal paste that’s used to try to dissipate the heat generated by the components. 

It’s not really going to work because it’s sealed in a plastic box with no effective heatsink. It’s a token gesture at best. At worst, it creates a mass that’ll slowly heat up and hold temperature because it’s got no way to get rid of it.

Behold the grey goo!

Adrian Kingsley-Hughes/ZDNET

Next to this goo was a bank of capacitors — the black cylinders in the photo — which were the cause of the failure. They’d clearly overheated, with three of them showing signs of bulging.

The problem!

Adrian Kingsley-Hughes/ZDNET

Well there’s the problem!

I also noticed that two of the components — bridge rectifiers that are used to turn AC mains into DC — have been fixed on an angle to make the touch a metal heatsink. It’s not really an effective way to cool down components.

The bottom line

Another “too good to be true” device bites the dust. It’s not the first one I’ve come across, and it won’t be the last.

Moral of the story here is that manufactures are using big number marketing — in this case 1,000W and masses of ports — to scalewash poor quality products. 

This might be a half-decent product if it was built to deliver 100W, but there’s no end of competition at that end of the market. Silkscreen “1,000W” on the outside, sprinkle in a few reviews that feel scripted and fake, and all of a sudden it’s interesting and exciting… right up until it blows up. 

Also: My top 7 laptop-bag essentials now, after decades of remote work

I know of no 1,000W charger. In fact, the 500W Ugreen Nexode is the highest-power charger that I’ve tested that’s legit. And the price is also legit — $250. 

But it’s built to deliver on what it promises and is packed with safety features, including “tip-over protection,” which cuts the output when the unit tips over and prevents it from falling on its side, where it can’t dissipate heat effectively. Now that’s an attention to safety that I like to see in a product that handles that much power. 

But if you want 1,000W of output, you’ll have to buy two and duct tape them together.





Source link