Tag: AuliaSays

AI voice detection and recognition are becoming more crucial

This Twitter thread shows how far along artificial voices have come. For those who are familiar with Steve Jobs’ voice, the voice in these recordings is almost indistinguishable from the original. When you listen to them, you can be forgiven to think that it’s actually Steve Jobs saying these words, never mind that he’s been gone for more than a decade.

The only catch is that because the training set must have been taken from the many recordings of his Apple keynote speeches and product announcements, they all sound like he’s reading from a script or making announcements. None of the sentences sound natural the way someone would speak if they were having a regular conversation or answering questions but that’s not too difficult to overcome. The tools to make adjustments to AI generated voices to sound more natural already exists.

Here’s another example. The YouTube channel Star Wars Comics have started to experiment with using generated voices to narrate some storyline’s from the Star Wars comic books to keep their audience up to date with what’s happening in the comics. In one video, they used James Earl Jones’ Darth Vader voice to say the lines in the pages of the comic book. Their latest video voiced a conversation between Emperor Palpatine and Darth Vader from another issue in the recent Darth Vader comic book series, both using the generated voices of their real actors.

As many in the comments noted, while their voices sound indistinguishable from the original, the speech patterns make it obvious that these were generated. That’s because the voices weren’t adjusted to the way a person would speak in a proper conversation given in the situation. Again, these are relatively trivial changes that one could make using their AI voice generators.

While these may be little more than fun projects for the curious minds, the day when someone can create entirely fabricated recordings to manipulate the public is already here. You can already create fake videos of a person saying things that they never actually said, now the voices sound even closer to the original.

When deepfake videos started popping up in 2020, people knew that this was going to be a significant problem. People are already easily fooled by fabricated articles or stories and this is just going to make it far more challenging for people to fact check and verify the validity of recordings.

All I can say for that is, brace for impact.

Just spent half the day helping my kid format her school paper on Word. I haven’t touched that software probably in more than a decade. Always worked on alternative apps like Pages, Notes, TextEdit, Docs, Bear, iA Writer, etc.

Talk about being overwhelmed. Had to hunt and peck each button and feature and search online for tips on how to do certain things.

Never got my parents to help me with any of those as a kid because they didn’t grow up or worked with computers and to this day they still don’t know how to operate one beyond using browsers and WhatsApp. Had to figure things out myself and it was practically troubleshooting which got me into tech support before journalism kicked in.

As someone who lived through his entire life surrounded by tech products and devices and worked as tech support for a significant period, this was kind of embarrassing but also challenging and humbling. It took much longer to complete than it ought to but got it done eventually.

Getting the brain to reconnect neurons that didn’t need to connect for a long time was quite an exercise. Pretty exciting, actually, but also exhausting.

The NY Times changed the headline of Jamelle Bouie’s opinion piece to be much less direct.

I haven’t been a Times subscriber in years and haven’t read anything from there for a while. The more I see this sort of behavior from their editorial team the more I’m convinced the Times is not worth the attention it’s getting. By no means it can be referred to as the paper of record.

The NY Times is a lot like Twitter. It’s an ongoing and escalating dumpster fire but it has so many eyeballs and attention people keep publishing and reading stuff there and triggering more complaints about them.

I get that the second headline is more clickbaity to draw in more curious readers (and page views) because the first headline will definitely be rejected by those who aren’t interested in the issue or will reject anything defending trans rights, but when you read Parker Molloy’s piece, it tells you that the headline change is much deeper than that. It’s more plausible that it’s a deliberate act against the trans community even if they try to argue it’s not the intention.

The NY Times have responded to the claims of being anti trans with a statement published by Nieman Lab. If ever there was a statement that nobody but the writer believes in it, this would be it.

Instagram founders Kevin Systrom and Mike Krieger have reteamed to launch Artifact, an AI driven personalized news feed that also lets you and your friends discuss about each story you share.

Casey Newton, who broke the news, calls it TikTok for text. From his description it sounds like AI driven Flipboard with chat.

As you can see from the screenshot registration is limited to US phone numbers for the time being. They’re clearly in testing mode and managing traffic. The whitelist, iPhone exclusive period, and pretty interface create a FOMO mindset among those who can’t get in yet.

It’s the stereotypical Silicon Valley product launch. Create a pretty app and generate widespread hype and FOMO by limiting access and riding on the news wave and public chatter.

From John Gruber’s take it’s not a promising start, maybe it’s far too early to judge. Right now it’s closer to a newsreader than a social app of any kind, and maybe that is the whole point and the conversation aspect is just a minor feature.

The things online stores do

I bought two boxes of masks that were supposed to be delivered yesterday. They didn’t arrive.

When I remembered this afternoon to check for the masks, my brother told me there was a package earlier on the bookshelf. It was a sealed black plastic bag. Ok, not boxes, which means I needed to check the app and see where my order was because the package felt like they could be clothes inside, not boxes of masks. When I checked the app it said they’ve been delivered. But there were no boxes delivered today, just the one plastic bag.

So I went to open the bag. And as I was opening it, I wondered if they took out the masks and repacked them without the boxes. I mean, okay, maybe they’re being sus about it or just didn’t want to send the boxes to make it a smaller package and save money.

I poured the contents out and sure enough, four small plastic bags full of masks. Suspicions almost confirmed. Right? 

Wrong.

Inside the bag there were the two boxes the masks were supposed to be inside of. But they were flattened. THEY WERE FLATTENED!!! These two boxes were looking freshly flat as if they just came out of the goddamn box production line! 

Geniuses, fucking geniuses, these people. Save money on the delivery and making them easier to transport by unboxing the masks and flattening the boxes.

Props to the sellers, man.

Speaking of third party Mastodon clients…

Checking out phanpy.social, it looks amazing. So clean, very app like, adjustable text size and interface via browser control. It’s still in development but it looks like it’s ready for regular use if you’re looking for a visually pleasing and functional web app for Mastodon. Best used on mobile or tablet, IMO.

All the boosted posts are grouped together in a carousel, threads get highlighted with an automated count, replies have a different shade, content warning, post visibility, text description, language options, attachments, and polls are available.

Posts are shown either as the normal stack of text blocks or separated as cards depending on your browser’s zoom level, and you can launch the post composition field into a different browser window for much larger and clearer view.

It has dark mode and post detail opens up in a new sheet over the timeline in the same window. Love it!

PS: Here’s an actively updated list of clients and resources for Mastodon

A Twitterrific for Mastodon?

I’m heartbroken that Twitterrific had to die and I don’t suppose Iconfactory plans to build a Mastodon app. Twitterrific developer Craig Hockenberry mentioned in his blog post that the new Mastodon apps are excellent to help with the transition but they ignore the scale and possibilities of ActivityPub. 

I can understand that. I mean, Mastodon is clearly the flagship ActivityPub client right now but its nature means you can see status updates or posts from people posting using other environments such as Pleroma, Misskey, CalcKey, Friendica, Pixelfed, PeerTube, Takahe, micro.blog, etc which are not Mastodon and not fully compatible with the Mastodon API, but they kind of work similarly in some ways yet at the same time quite differently as well.

Should one build an all encompassing Activity Pub app? I don’t think so because it would end up as a big confusing hot mess. Apps optimized for individual services are still the way to go  in my opinion. I’m still hopeful that there’s an Iconfactory take that can accommodate the Fediverse because the team has always managed to come up with innovative designs and user experience. They’ve never been afraid of experimenting with different approaches to the same problem.

Maybe they have something in mind already or secretly have been building something, maybe they haven’t. Given the response they gave to being abruptly shut out of Twitter, maybe they’re going to take some time before deciding on their next step.

Right now there’s already plenty of upcoming third party Mastodon apps to choose from for iOS even if they’re still in TestFlight, not to mention existing ones as well as web apps that can work across platforms, but a buffet of choices never stopped new ones from appearing.

The death of third party Twitter apps

On Friday, 13 January 2023, Twitter turned off access to most third party Twitter clients. People who used apps like Tweetbot, Twitterrific, and Echofon were suddenly shut out of their apps, faced with a screen that told them there’s an authentication issue. 

It’s not the first time third party apps were given the shaft by the company but this time there were no warnings, no official statements acknowledging the matter, no explanations. It’s as if millions of voices were suddenly silenced.

While there had been internal discussions about the ban, it wasn’t until almost a week later that there was an official word and even then there was no public communication that third party apps are no longer allowed. Engadget found out by diving into Twitter’s developer terms which were only just updated yesterday.

Twitterrific developer Craig Hockenberry was rightly incensed about the move, and his Iconfactory colleague Sean Heber wrote about having to shut down the app after 16 years. It’s absolutely clear that Twitter is no longer the same company as it was a year ago or even six months ago.

Unlike companies such as Instagram, Snapchat, YouTube, or TikTok, Twitter had always been reliant on third party apps and solutions even near the beginning, which was why the 2012 API restrictions were met with massive backlash to which Twitter then relented. This time, though, no such luck.

The LVCC Loop is a Glorified Theme Park Ride

This is someone’s idea of a revolutionary rapid transport that’s supposed to beat the traffic. Sure, it’s currently only a prototype, just a showcase of an idea, but it looks very much as if the future of transport is a Disneyland ride or a glorified Uber pool. Other cities have much more efficient and safer versions called a bus route, a metro, or MRT.

You may have read about last year’s Tesla jam in the Las Vegas Convention Center Loop so there’s no question that it can happen again.