If you’re paying for something with your data, then that makes data a currency.
So it seems New Zealand is at it again when it comes to leadership on the world stage. The government has released an ‘algorithm charter’ that aims to tackle the lack of transparency and potential bias in automated systems.
This is a long time coming, and other countries need to pay attention.
Whilst the document is a great start, its language is somewhat vague and aspirational - perhaps intentionally so, given it is early days for this kind of scheme.
As AI and automated systems become evermore present in our daily lives, with the potential to decide your next job or even the next US president - I don’t think it’s unreasonable to expect companies to be more transparent about how their AI works.
Just as we expect companies to file accounts and keep records, we should ask companies to answer a few basic questions about how their algorithms work. This will give consumers confidence that they’re not subject to unfair bias when decisions that affect them are made by AI.
To make AI more transparent, I’d ask companies for:
A high level description of the type of algorithms in use - most machine learning algorithms are not written from scratch - so just tell us for example whether the algorithm is a Logistic Regression or a Transformer Network etc.
The source of the data used to train the model - did the company collect it themselves, if so, where from? Or did they buy it from a 3rd party, if so, who, and where did they get it from?
The features used to train the model - knowing for example that ‘sex’ or ‘age’ are fed into the model would tell us where to look for potential bias.
The methods, and datasets used to validate their model’s effectiveness - this is critical - a company without a good validation strategy is obviously going to have unconscious bias in their systems.
I’ve picked the questions carefully as I think companies have a right to maintain their trade secrets. The data itself can remain secret, as should the various parameters and other specifics about the algorithm.
While I’m not expecting the average consumer to understand the different between machine learning algorithms and the consequences of different testing strategies, having this information in the public domain would enable journalists and the tech community to better hold big tech companies to account.
The privacy invasion is the result of the apps repeatedly reading any text that happens to reside in clipboards, which computers and other devices use to store data that has been cut or copied from things like password managers and email programs
I love that Apple have started to notify users when an app reads the clipboard without the user explicitly pressing ‘Paste’, however I worry that adding yet more warnings and notifications will result in users becoming fatigued and eventually ignoring them.
Instead, why not allow apps that need to store sensitive data on the clipboard (LastPass, for example) to set a specific data type indicating that the information should be classed as sensitive data - much like they can set the data type to text or image at the moment. In order to paste this “secure object”, users would have to press the system ‘Paste’ button - any app just reading the clipboard by itself would just see it as empty.
This of course wouldn’t solve for cases where data is not known to be sensitive - how many people store their passwords as a simple note instead of in a password manager, for example? Still, fewer notifications and stopping any app from reading passwords from the clipboard might be the right balance.
Ofcom also found most people who use an address provided by their broadband company are former BT customers. BT says if customers want to switch provider but keep hold of their old BT email address they’re charged £7.50 per month to be able to access and use their account like they used to, including accessing it using an app.
Reading beyond the BBC headline, it turns out that BT are in fact allowing former customers to access their old email account for free, but this only works using a web browser, presumably this is because it’s funded by advertisements - I can’t see any other reason for this limitation. It is only for continued access via POP/IMAP that BT demand payment for.
Even so, £7.50 a month seems extortionate considering you can get a full Microsoft Exchange account for ~£4.50 a month including VAT.
This kind of exploitative pricing is in poor taste, especially considering that in many cases - the type of customer who still uses their ISP for email is most likely still running Outlook Express - as it was configured in 2007 - and so any change will be more confusing and disorientating than it would for someone who knows how to setup a Gmail account and configure email forwarding (BT do offer forwarding, but I’ve no idea if this is covered by the free option).
We solved this problem with mobile operators by having rules that mean that as a consumer, you have the right to take your phone number with you when you switch to another operator. Email is unfortunately not regulated in the same way, and in fact its architecture makes it impossible for an email address with the domain “btinternet.com” to be be routed to a server owned by another ISP. Instead, I’d like to see the option to forward emails for free to another email address for a set period of time (say 2 years). If major ISPs and popular webmail providers could somehow automate this process, even better.
Though email is arguably less relevant today than it was 15 years ago, I would still strongly advocate buying your own domain name (around £15 a year) and then setting it up either to forward to a free email account, or even better, paying for proper email with the domain (around £2/month with FastHosts). For those who fancy a technical challenge, for around £4/month you can setup a lightweight Debian VM on somewhere like Linode and run your own mail server - this gives you the benefit of being able to setup unlimited email addressed with your domain - though it’s not for the fainthearted.
I’ve not yet paid for a Hey.com subscription as I’m a strong believer in owning your own domain name for email, but I have to admit I that find their style of business a welcome change to the norm and reminiscent of what I can only imagine the counter-culture inspired Apple must have been like in the mid 1980s.
Enter GMass, a plugin for Gmail that adds spy-pixel tracking, amongst a grab bag of other stuff. They hadn’t been on our original list of 50+ services we name’n’shame, but thanks to a new blog post where they brag about defeating protections that recipients might take to defend themselves, they came onto our radar.
It turns out the makers of this Gmail plugin have used the unimaginative steps of using URLs with no obvious parameters (we called them “friendly URLs” in 2006), encryption of those parameters (not so friendly), and setting up a CNAME record so senders can appear to be hosting the tracking images on their own server, all to try and avoid being tracked by Hey.com (the irony, right?).
Whatever your take on tracking pixels, I just think it’s great to see a company saying it how they see it and taking a strong stance. Personally, I’ve long set my email client never to automatically display images - not so much because of tracking pixels from companies like GMass, but avoid confirming my address whenever I accidentally open up spam.
The Maximum PC site is still online to this day, as it was when it was abandoned in 2001
On the most recent episode of ATP, John Siracusa referenced once of his early Mac OS X reviews. This prompted me to revisit his review of the Mac OS X public beta, as the evolution of computer interfaces is something that I find fascinating.
In one of his screenshots, he shows the Ars Technica home page. At the top of the page is a banner informing readers that Ars is part of the “Maximum PC Network”. I immediately recalled this web site, as it was closely related to many other publications I was into at the time, including .net magazine. So, I decided to see what became of the Maximum PC Network by visiting its site in 2020. I expected to find a domain for sale, or it redirecting to a parent company - but no, to my surprise, maximumpc.co.uk is still online, and is still showing the date of when it was presumably last updated - 26th February 2001. Complete with references to em@il and asking whether Windows XP is worth upgrading to, the site is straight out of 2001, but amazingly - still online. Some server, somewhere in 2020 has been left running sine 2001.
I’d love to know the story of how this has happened - how does someone forget a web site existed, but continue to pay for its hosting – even going as far as renewing the SSL certificate each year? That said, I think it’s great and I wish more web sites could stay online once they are abandoned.
To reduce battery aging, AirPods learn from your daily charging routine so they can wait to finish charging past 80% until you need to use them
I have long speculated that the reason AirPods are much more disposable when compared to other Bluetooth devices is that fact that their poor batteries are pretty much always in constant use. They are always either in the case charging, or in your ears discharging.
I own a pair of Plantronics wireless headphones from 2015 that have much better battery life than my AirPods which were purchased in 2016. I know there are many factors at work but the speed at which the batteries on AirPods decline is staggering. I suspect (and hope) this update will dramatically increase the longevity of AirPods.
I recently came across the sad news that .net - the internet magazine published its last issue in April 2020.
.net magazine was hugely influential to me as a teenager, capturing the excitement of the Internet in the late 90s. Some of my earliest memories of using the Internet revolve around .net magazine. Within its pages I learnt many things about how the Internet works; how to configure domain names, what emoticons meant, tips to avoid spam, what a flamewar was and how to solve problems with Netscape and Outlook Express, to name a few. The first issue I remember buying (issue 54 if I recall correctly, in late 1998) even had an interview with David Bowie, talking about amongst other things, his recently launched Internet Service Provider (ISP), BowieNet. Bowie was incredibly forward thinking, however music artists running their own ISPs never did take off (though I seem to recall some who’ve been successful at making headphones). [Update: it was issue 55, and I’ve scanned the Bowie interview for anyone interested]
Even now, 22 years later, I can recall the excitement of installing the Internet Explorer 5 beta from the cover CD and marvelling at how it dramatically changed the way Windows 98 operated. Strangely, I have a memory of disconnecting from the Internet, in order to phone a hotline to get a code that would let me use one of the applications included on the cover CD, “Starfish Internet Sidekick”. The code was “MANAGER”. It’s weird how such memories stick with us. My first look at Mac OS X was from a preview in .net magazine, which touted its photorealistic icons and UNIX underpinnings - I never thought back then that one day in the year 2020 I’d be sitting at a laptop writing this, running the same OS, still using essentially the same UI paradigm, only with far less photorealistic icons.
As a teenager who wasn’t really into books, .net magazine taught me more than how to use the Internet - it helped me to increase my reading age, develop an inner voice and writing style that is still with me today.
What I looked forward to each month the most however, was waiting to see if any of my forum posts had been picked for inclusion in the ‘penny arcade’ section - where four or five of the funniest posts from the .net forums were featured in the magazine. I don’t think my posts contained enough witticisms to make it into the magazine, but just the possibility was exciting.
The .net forums were a magical community of likeminded geeks, mostly in their teens, who just loved the Internet. These days, nobody “loves the Internet” as we did then. Today that would be like having “electricity” as a hobby. Yet back from around 1999 to 2001, the .net forums were where I spent a considerable amount of time socialising (though at the age of 15, my parents didn’t think of it as being very social). I remember posting a message to the Futurenet NNTP server through Outlook Express and within minutes getting a reply. It was like Reddit, but far more humane, and because there were only about 20 or so regulars, it felt like a real community. I remember the Microsoft .NET announcement - how could Microsoft steal the name of our beloved magazine? Waiting for the world to end as Nostradamus predicted in 1999, checking that the internet still worked on January 1st 2000, and posting a link to my latest Geocities creation to the reader’s site newsgroup - eagerly awaiting feedback from fellow readers - all on the .net forums. The community faltered slightly when Future Publishing closed down the NNTP servers and moved the forums to web-based software. Thankfully many of forum regulars found a new home on IRC in #netmag on the Blitzed network. There I would spend whole weekends chatting to my friends on IRC, in-between TFC sessions. It wasn’t long after that we built a site called “DFNET - The .net reader’s site network” which further reinforced the sense of community between readers.
That idea of having your own website and even forming a community of sites is something I miss from today’s Internet. The web is now so centralised. How many ISPs even bother offering free webspace today? Back then, out of the entire online population, there was a disproportionate percentage of computer enthusiasts online, and so learning HTML and how FTP works was a challenge they would gladly take on. People created web pages about cooking, Half Life and the X-Files. It’s easy to look back and laugh at the animated Gifs, the “best viewed in Internet Explorer 4” badges and the “Sign my Guestbook” links - but a part of me mourns the loss of this amateur enthusiasm - the web is now so much more polished and corporate, and “user generated content” mostly resides on major platforms like Twitter, Facebook and IMDB. I would love for Internet providers and OS manufactures to activity encourage people to make their own websites again (remember when Windows used to include FrontPage Express?), but I think that is sadly a pipe-dream and the reality is most people aren’t interested. With modern broadband connections, personal websites could even be hosted at home - unrealistic - but part of me is nostgltigc for the dream of an Internet where everyone is an equal player.
I must have stopped subscribing to .net sometime in 2002. By then, the Internet seemed less interesting to me. Perhaps due to the .com crash or because I was 17, now at college with a Saturday job, and for some reason didn’t have time for it any more. Aged 17, the gap between 1999 and 2002 felt like a lifetime. Now, aged 35, three years ago seems like yesterday. The magazine eventually morphed into a web design magazine, focusing less on Internet culture and more on design trends and technical implementation. I picked it up again in 2009 for a year or so but as I became more professionally interested in server side technologies the design centric content wasn’t as appealing to me (I am no designer!).
Although I hadn’t read it for a many years, hearing of its demise did make me sad that a part of my childhood is no longer. I’m gutted for those who worked at the magazine and may now be out of a job. Wouldn’t it be a nice gesture if Future Publishing produced PDFs of the magazine archives (they surely must exist, right?) and put them online for historical record? I could genuinely lose hours reading Internet articles from 2000.
With all the kerfuffle over Hey.com - an exciting new email service from the makers of Basecamp, it’s important to remember that if it takes off, Hey.com will be in a similar position to Apple. Groundskeeper of the walled garden.
While the service appears to be a refreshingly original rethink on how email should work (I’ve not signed up myself), it appears to be designed to lock users in. There is no way to use an external email client, and right now there is no support for custom domains (despite being targeted at business users with a $99/year charge - they say its coming). You’ll also have to continue to pay if you decide to move to a different provider but want to keep your Hey.com address forwarding. I see no reason why the features advertised (great as they are) couldn’t be part of an email client designed by Hey.com - using a combination of cloud services and native clients, Hey could provide a powerful front-end to any open email service such as Outlook.com or Gmail which provide access via IMAP.
Instead it seems they want to lock customers in, which goes against the open, standards based approach that has made email so successful so far. Putting aside the fact Google or Microsoft could easily copy any of these features, and probably will, this is a service I’ll be skipping.
Update: thanks to Shauny for informing me that Hey.com does in fact offer free forwarding for life for customers who pay for a year’s service.
As someone who is fascinated by technology and its societal impact, I often think back to the first piece of technology I remember owning as a child: A Casio Telememo 50 digital watch (photo from the Argos catalog I would have ordered it from).
Back in 1996, the idea of storing phone numbers on your wrist was a niche that probably only appealed to geeks like me. Now, in 2020 millions of people walk around with smartwatches and don’t even realise they have their contacts synchronised to their wrist.
I can’t find the exact model I owned as a child, the photograph above is the nearest I’ve been able to find on Amazon, and is unfortunately not as good.
Here’s what I remember being so cool about my first smartwatch in 1996:
The blue illuminator backlight looked so cool in the dark. I was always worried about using it however, as it drained the battery big time, and aged 11 new batteries aren’t easy to come by. Best thing though? It flashed when the alarm went off.
The hourly time signal - during school assemblies, from about 5 minutes to the hour until about 5 minutes after, there was a chorus of “beep beep” sounds as everyone’s Casio watches went off. Mine of course would always be perfectly always accurate as I would set it to the time shown on Ceefax (much cheaper than the other option of phoning the speaking clock).
World time - At 11 years old, I had no need to know the time in other countries but it was so cool being able to tell my friends, uncles, and aunts what time it was in New York when I visited them.
Scheduled alarms - Long before anyone would dream of being able to set a calendar appointment on their phone (an Americanism - back then we’d have called it a ‘diary entry’), my watch would let you set an alarm for a specific time, on a specific date. Whether you’d hear it or not was a different matter. I didn’t have many events happening aged 11, but I do remember setting it for my occasional hospital appointments.
Water resistant - I wore this thing swimming, in the shower, in the sea. Quite a novelty at the time to be able to have an alarm go off to remind you its time to get out of the pool.
If my 11 year old self had any sense, he would have kept this watch, but alas it is sadly most likely to be landfill somewhere now. I’ve found a similar one on Amazon, but it has far fewer features and doesn’t look anywhere near as cool.
“Microsoft ‘to replace journalists with robots’” from BBC News
Microsoft is to replace dozens of contract journalists on its MSN website and use automated systems to select news stories, US and UK media report.
I’d love to know why somanynewsorganisations are using the term “robots” to describe algorithmic selection of content. Google has been doing this since 2002 with Google News, so I’m surprised that this is even news (though I’m of course sad for those who are now out of a job). I’ve always thought the term “robot” to be a pointless one. We don’t call our washing machines robots, so why would we call any other machine that isn’t anthropomorphised a robot?
I can’t help thinking this is also being used as an excuse to remind people MSN News still exists.
