I was trying to figure out why people still use Axios, when the built-in fetch works just fine. Is it because people are still sending XML requests?

Removed by moderator

.

.

Gotcha! This comment has not been removed… Probably.

Of course they do. They are extremely impartial on the matter and I trust their judgment.

I agree it would be good to have third party integrity checks to not require Google Services etc. as part of the chain.

In GrapheneOS, many Google Play integrity check pass, but payments still do not work. You are notified when an app uses the integrity API, but probably only because they have spent a bunch of work sandboxing Play Services. This is what you see when you look at those details:

I guess the obvious problem is that so many apps rely on Google Services, such as for payments, opening the store, checking for integrity etc. On stock android, you can’t pick and choose these services separately or use third party ones, unlike using a third party keyboard, for example. Everything is one big proprietary, data guzzling lump.

Let me guess:

It’s because all the money goes to billionaires.

Edit: Pretty much what it says. It’s more detailed than that but yeah. Labourers get less, more value is attributed to capital (buildings, land) and collected by the rich.

Our Pixel ⚒️🎵

The one I hear is good is NVDA by NVAccess, but it doesn’t have AI.

Website: https://www.nvaccess.org/about-nvda/

Source code: https://github.com/nvaccess/nvda

I would be fairly surprised if you found an open source screen reader that had AI built-in.

It would likely have to run locally if it was open source.

Typically companies don’t make their stuff open source, and non-profits are unlikely to host an AI for you because of the cost. It’s unlikely to run locally as the cost for that model to run and download size may make it unusable or impractical for some hardware. Typically screen readers need to be accessible to everyone, and therefore need to run on very old and / or cheap hardware.

If you are speaking about the UK’s Labour party, in it’s current form I would describe it as further left than the US democrats, but further right than the actual left wing parties like the Greens, Your Party and Bernie Sanders (and any other tiny socialist groups).

To me, that means center-right generally. Fully right on certain issues such as immigrants and trans rights.

Before Kier Starmer and Labour Together, I think they were center-left. Many people within the party are still trying to do good things and vote against Labour’s worst policies.

The reason I hate Labour even more than the Conservatives is at least with the Conservatives you knew what you were getting. Here it feels like betrayal.

Given that only 22 noncitizens over the course of like 10 years (can’t remember the exact statistic) tried to vote, the way to get the best accuracy would be to just hard code the answer to “yes, they can vote”. Then you get over 99.99% accuracy.

Mental gymnastics on full display

Another interesting one. These extensions are all related:

I’ve gone through the list a bit and out of the most popular ones that spied on you, most were adblocks, coupon finders or AI Chatbots.

Some notable extensions:

• Stylish. A theming extension, I used to use this back in the day!
• Smarty. Some sort of coupon code thing like Honey
• Video Ad Blocker Plus for YouTube™
• Video Downloader PLUS
• Karma - Another coupon thing
• Audio editor online Audacity. Some sort of web-based Audacity clone?
• GIMP online - Same sort of thing as above with GIMP
• Ground News Bias Checker - To be fair it probably makes sense this one sends the URL you are visiting, as it’s purpose is to look up the bias of the publication you are looking at.

Worth a read regardless.

Great work to the investigators here. I’m going to comb through this list a little. See what things stand out.

The point is that I’ve seen several comments on other posts about this vulnerability, and in the body of this one, saying that Notepad is bloated and terrible now.

I’m offering a counterpoint that this is not necessarily bloat. It’s debatable that this is the right tool to have this feature, but it can be a useful feature.

I’m fine with Markdown support, but I wish MS got the message about Copilot being unwanted. Not sure if they’ve added it to Notepad or not at this stage, but given all the places they’ve crammed it into I wouldn’t be surprised.

To be fair, markdown is a very cool standard.

While I don’t know if it really makes sense for Notepad to be anything other than a plain-text editor, there are better tools for that, supporting markdown is kind of nice.

This means you have support for it on fresh Windows installs, which could be good for virtual machines. That said, Markdown is intrinsically pretty readable without formatting anyway.

It’s a shame they flubbed the implementation though…

Or percentages

Related video

https://youtu.be/uDkyP37JgY0

hunting for alternatives

I’ve heard people mention Matrix, but I’ve not tried it yet.

Hmmm…

As the article correctly states, machine learning (“AI” is a misnomer that has stuck imo) has been used successfully for decades in medicine.

Machine learning is inherently about spotting patterns and inferring from them. The problem, I think, is two-fold:

1. There are more “AI” products than ever, not all companies build it in responsibly and it’s difficult for regulators to keep up with them.

The gutting of these regulatory agencies by the current US administration does not help ofc, but many of them were already severely undermanned.

2. As AI is normalised, some doctors will put too much trust in these systems.

This isn’t helped by the fact that the makers of these products are likely to exaggerate the capabilities of their products. This may be reflected in the products themselves, where they may not properly communicate the degree of certainty of a diagnosis / conclusion (e.g. “30% certainty this lesion is cancerous”)

AI radios

Looks inside

No AI

Jokes aside, this is a common thing in tech / software at the moment.

You can make fantastic software and systems, but unless you slap an AI label on it, big companies and organisations will not want to pay for it, or will pass you over for a product that says it has it, even if it’s dogshit.

AI (or, more accurately, machine learning) can bring value, but so can a lot of other features.