#Tech Criticism

[ahmetasabanci]

Formulated Criticism and Flattened Futures

There’s a trend I’ve been watching for some time about how the critical approach to technology is changing, and becoming more and more boring and formulated as the time passes. I started to write a short blog post about that one but things started to connect in my head in an unexpected way and realized that this is not only about technology but also about how people are looking at future in…

[ivygorgon]

Well shit. I walk back what I said. that works for my spouse's phone too. Wth. Thanks for solving that for me.

yknow what would be a fucked up phone feature

[boredtechnologist]

Microsoft Copilot Vision: Privacy in Name Only

The following analysis is an independent critique intended for informational and internal review purposes only. It does not reflect the views of Microsoft Corporation and is not affiliated with or endorsed by Microsoft in any way. This content is based on publicly available information and does not assert the presence of any specific legal violation. Readers are encouraged to consult legal counsel or official policy documentation for compliance decisions or interpretations. Microsoft has introduced Copilot Vision as an enhancement to the Windows user experience, encouraging users to share their desktop and application windows with an AI assistant that can analyze content in real time. The messaging frames this as an opt-in tool for productivity and support, but behind the polished interface and helpful tone lies a far more invasive mechanism. Copilot Vision extends deep visibility into a user's screen, with few meaningful boundaries, little transparency, and even less accountability.

Despite lengthy privacy statements published by Microsoft, the actual safeguards for screen data shared with Copilot remain vague at best. The policies speak at length about privacy values and responsible AI. What they do not clarify is whether your screen content, once shared, is processed locally or remotely. They also fail to define what data is retained, how long it is kept, who within Microsoft may access it, or whether any of it is used to train future models.

The privacy language makes frequent use of ambiguous phrasing like "we may use your data to improve our services" or "data is used to provide a better experience." That wording is broad enough to justify nearly any use case, including behavioral analysis and long-term retention for AI model improvement. The result is a kind of opt-in surveillance that the average user does not fully understand and cannot easily control.

For example, when Microsoft says Copilot can "see what you see," it means exactly that. Anything visible on your screen, from customer records to financial dashboards to personal photos or encrypted communications, is made available to their system. There are no clear visual indicators when this is happening beyond a small icon. There is no automatic redaction of sensitive fields. There is no evidence that the content is ever processed in a zero-trust model or confined to temporary, non-persistent memory.

Even more concerning is the language used in the Enterprise and Developer Products section of the privacy statement. It outlines broad allowances for data use in support of Microsoft's business operations, ranging from troubleshooting to workforce development. There is no guarantee that data shared through Copilot Vision is exempt from this. Enterprise customers may believe their data is protected by contract, but those protections only apply if negotiated explicitly. Most users are unaware of these distinctions and assume privacy controls are enforced by default. They are not.

The consumer version of Copilot, including its Vision feature, does not provide enterprise-grade controls unless specifically enabled through Microsoft’s commercial data protection offerings. However, even with those in place, the boundaries remain blurry. Microsoft confirms that both automated and manual methods may be used to process your data, including direct human review of AI outputs. That effectively gives employees or contractors the ability to view data collected through this tool. While Microsoft claims to follow responsible AI principles, the implementation of those principles is difficult to verify and rarely exposed to third-party audit.

The most telling detail comes from the privacy section on children and education. Microsoft goes out of its way to assure parents that student data will not be used for advertising or behavioral profiling. Adults, however, receive no such promise. For everyone else, the data is subject to Microsoft’s full range of operational, analytical, and marketing use cases.

The key problem is not that Microsoft has built a system capable of watching your screen. It is that they have built it with minimal restriction, cloaked it in helpful language, and buried its implications under hundreds of paragraphs of policy text. Most users will never read that far. Even fewer will understand how much of their working environment they have just handed over.

Any claim that Copilot Vision operates within user consent ignores the reality that most consent is neither informed nor reversible. Once content is seen by the system, there is no button that makes it unseen. Microsoft may offer options to view or delete portions of your data through its dashboard, but that applies only to specific categories. There is no assurance that full-screen content shared with Copilot can be reviewed or purged, nor is there any audit trail made available to the end user.

In practical terms, Microsoft is inviting users to broadcast their digital environment to a remote system that operates according to complex, shifting policies. These policies are not easy to find, are not written in plain language, and often allow Microsoft to use the data in ways that serve its commercial interests more than the user’s.

Copilot Vision represents a new level of access. It is not a benign helper waiting to respond to questions. It is an AI system that watches and learns. The real concern is not just what it can do now, but what it will do next, and whether users will have any say in the matter. Privacy cannot be preserved through long documents alone. It requires structural limitations, transparent enforcement, and the willingness to place user protection above platform growth. Microsoft’s current approach does not meet that standard.

If an organization values confidentiality, compliance, or basic user trust, this feature should be considered high risk. The benefits of convenience do not outweigh the exposure it creates. This document is provided for critical analysis and educational discussion. It should not be construed as legal advice or an accusation of wrongdoing. All trademarks and product names mentioned are the property of their respective owners. Use of this material is subject to fair use principles for commentary and review. For specific guidance regarding data privacy and security practices, consult a qualified professional. Link:

Copilot Vision on Windows with Highlights is now available in the U.S.  | Microsoft Copilot Blog

[queen-mabs-revenge]

The rise of neoliberalism as a dominant political paradigm in the 1970s and the emergence of platform capitalism in the 21st century resulted in new areas of employment and a decrease in the relative numbers of workers employed in traditional working-class jobs. Data work, that is, the labor that goes into producing data for so-called “intelligent” systems, not only brings with it unprecedented forms of exploitation but, crucially, possibilities for organization and resistance in globally networked economies. 

The Data Workers’ Inquiry is a community-based research project in which data workers join us as community researchers to lead their own inquiry in their respective workplaces. The community researchers guide the direction of the research, such that it is oriented towards their needs and goals of building workplace power but supported by formally trained qualitative researchers.

We adapt Marx’s 1880 Workers’ Inquiry to the phenomenon of data workers who are both essential for contemporary AI applications yet precariously employed—if at all—and politically dispersed.

[bravecrab]

I've previously recommended a bunch of podcast episodes covering AI hype criticism, and Emily M. Bender has featured in quite a few, so I want to promote the podcast she's doing with Alex Hanna, Mystery AI Hype Theater 3000.

Spotify

Apple

[fehck]

[ecrivainsolitaire]

A summary of the Chinese AI situation, for the uninitiated.

These are scores on different tests that are designed to see how accurate a Large Language Model is in different areas of knowledge. As you know, OpenAI is partners with Microsoft, so these are the scores for ChatGPT and Copilot. DeepSeek is the Chinese model that got released a week ago. The rest are open source models, which means everyone is free to use them as they please, including the average Tumblr user. You can run them from the servers of the companies that made them for a subscription, or you can download them to install locally on your own computer. However, the computer requirements so far are so high that only a few people currently have the machines at home required to run it.

Yes, this is why AI uses so much electricity. As with any technology, the early models are highly inefficient. Think how a Ford T needed a long chimney to get rid of a ton of black smoke, which was unused petrol. Over the next hundred years combustion engines have become much more efficient, but they still waste a lot of energy, which is why we need to move towards renewable electricity and sustainable battery technology. But that's a topic for another day.

As you can see from the scores, are around the same accuracy. These tests are in constant evolution as well: as soon as they start becoming obsolete, new ones are released to adjust for a more complicated benchmark. The new models are trained using different machine learning techniques, and in theory, the goal is to make them faster and more efficient so they can operate with less power, much like modern cars use way less energy and produce far less pollution than the Ford T.

However, computing power requirements kept scaling up, so you're either tied to the subscription or forced to pay for a latest gen PC, which is why NVIDIA, AMD, Intel and all the other chip companies were investing hard on much more powerful GPUs and NPUs. For now all we need to know about those is that they're expensive, use a lot of electricity, and are required to operate the bots at superhuman speed (literally, all those clickbait posts about how AI was secretly 150 Indian men in a trenchcoat were nonsense).

Because the chip companies have been working hard on making big, bulky, powerful chips with massive fans that are up to the task, their stock value was skyrocketing, and because of that, everyone started to use AI as a marketing trend. See, marketing people are not smart, and they don't understand computers. Furthermore, marketing people think you're stupid, and because of their biased frame of reference, they think you're two snores short of brain-dead. The entire point of their existence is to turn tall tales into capital. So they don't know or care about what AI is or what it's useful for. They just saw Number Go Up for the AI companies and decided "AI is a magic cow we can milk forever". Sometimes it's not even AI, they just use old software and rebrand it, much like convection ovens became air fryers.

Well, now we're up to date. So what did DepSeek release that did a 9/11 on NVIDIA stock prices and popped the AI bubble?

Oh, I would not want to be an OpenAI investor right now either. A token is basically one Unicode character (it's more complicated than that but you can google that on your own time). That cost means you could input the entire works of Stephen King for under a dollar. Yes, including electricity costs. DeepSeek has jumped from a Ford T to a Subaru in terms of pollution and water use.

The issue here is not only input cost, though; all that data needs to be available live, in the RAM; this is why you need powerful, expensive chips in order to-

Holy shit.

I'm not going to detail all the numbers but I'm going to focus on the chip required: an RTX 3090. This is a gaming GPU that came out as the top of the line, the stuff South Korean LoL players buy…

Or they did, in September 2020. We're currently two generations ahead, on the RTX 5090.

What this is telling all those people who just sold their high-end gaming rig to be able to afford a machine that can run the latest ChatGPT locally, is that the person who bought it from them can run something basically just as powerful on their old one.

Which means that all those GPUs and NPUs that are being made, and all those deals Microsoft signed to have control of the AI market, have just lost a lot of their pulling power.

Well, I mean, the ChatGPT subscription is 20 bucks a month, surely the Chinese are charging a fortune for-

Oh. So it's free for everyone and you can use it or modify it however you want, no subscription, no unpayable electric bill, no handing Microsoft all of your private data, you can just run it on a relatively inexpensive PC. You could probably even run it on a phone in a couple years.

Oh, if only China had massive phone manufacturers that have a foot in the market everywhere except the US because the president had a tantrum eight years ago.

So… yeah, China just destabilised the global economy with a torrent file.

[ganondoodle]

wont have time to work on anything for a few days probably, so uh, since i love reading comments/tags of people sharing their experiences- as part of the preparation for the totk rant script i got another question to ask :3

if you dont like tears of the kingdom, was there a moment that "broke" you, as in, the moment you knew this game is worse than you thought/hoped, and if so what was it?

personally, while i was suspicious after seeing its last trailer, i told myself its just me again and i kept up my hopes for a long time into my playthrough- its hard to point to a specific point since it was a growing feeling of something being off, things didnt make sense and i ever so more wondered how they would pull this all together (they didnt)- i do think the moment i stopped being in denial about it was when i found the shrine of life, the beginning of botw, and found .. nothing, a dingy cave practically licked clean of any traces of the shiekah tech like it never existed, instead of the medical bed a pathetic puddle of water that healed you, no one caring at all, like it actually never happened- i felt like the game pointed and laughed at me for caring about botw, pretty sure i was struggeling to keep it together on stream bc it forced me to realize this game truly is everything i hoped it wouldnt be, even if that sounds a little weird, at that time zelda and especially botw was so much more important to me, a passion for the franchise this game really did end up killing.

[raymurata]

Just scroll (or go ahead and block my #dav critical tag) if you don't wanna read me whining abt Bellara's Archive choice again, but I'm not done with the salt.

What bugs me isn't that the choice to destroy the archive exists, it's that the game frames it, through its UI (which is the closest thing we have to a nonbiased narrator in this medium), as equally weighted against the opposite choice.

If they had worded it like this:

"Free the archive (the knowledge will be lost)" x "Keep the archive (the knowledge will be kept)"

With no extra commentary, then that would be better. If you got to be openly racist against the Dalish or openly in favor of the Dalish, period. Just like in previous games.

Bellara says "(the archive would help us) get back what made us who we were," and "With it, we could be that again."

Which is funny because... People don't study history to return to the past. It's fine if Bellara is idealistic and saying whatever unrealistic, grandiose dreams she and Cyrian had, but the Dalish would never (could never) become like the ancient elves again. For starters there is a Veil now. So what it would in fact do is help them understand where they come from, what they've been through, and trace the changes in their culture.

Of course our modern historians and scientists have tried to reclaim lost technology, too. They've figured out how the Romans made their extra sturdy concrete, and scientists in Brazil have long been trying to replicate the extra fertile Terra Preta from indigenous peoples that lived in the Amazon basin, and several South American historians would love to know how exactly the Inca used the quipu as a writing system aside from counting tool, etc... And that's super cool!!! And maybe (but that's a big maybe) the Archive could give the Dalish a technological edge to carve a corner of the world for themselves without the constant struggle with Tevinter trying to enslave them or Andrastians trying to subjugate them.

But I personally don't think anyone's reading Aztec accounts of human sacrifices to replicate the same practices in modern cults, or that there is an army out there utilizing Roman decimation as a method of discipline. We're using different horrific methods of control now, lol.

But let's say a modern general does decide that the best way to punish a battalion for one man's insurgence is to force every group of ten soldiers to violently murder the 10th man.... Do you really think that the fault would lie with the historian who unearthed that information and put it on Wikipedia? Or the insane general that decided to do this? Would modern morality and laws allow for that punishment to be executed? Do you think that the existence of that article online is inherently dangerous and controversial, and that it should be taken down? Do you think this general would have been a good and non-violent general if he hadn't ever read about Decimation? Or is it clear to you that violence and ingenuity are both inherent to mortals as a whole and can't be so easily blamed on the spread of knowledge?

Because it's not clear to DAV. The game (not Bellara, not Varric) words it very unambiguously as a dichotomy: The only safe way to deal with this Archive is to destroy it. Keeping it is inherently dangerous because the knowledge could fall in "the wrong hands."

What Bellara says is "Cyrian is gone because of what that thing knew," and "what about the bad side, the other things we did?" and "We stole the dwarves' dreams."

Again, she gets to say whatever she wants because she's a character and she's an anxious, idealistic mess. Love her for it. I like that she feels guilt here too because she has been established (through her way of dealing with Cyrian's first death) as someone who takes the blame for mistakes she didn't even commit (She certainly isn't responsible for Solas' actions). She's someone who drives herself sick cooking up the most horrific scenarios in her mind, and she's so compassionate she can't stand the thought of being the one perpetrating violence against innocents. Her misplaced guilt and dread are the emotions that lead her to consider destroying the Archive.

But no matter how guilty a young german may feel about the holocaust, destroying knowledge about gas chambers is not what will prevent other genocides from happening around the world. Individual guilt is barely productive.

Furthermore, Corinne Bursche says that DAV gives you a choice between "destroying" or "sharing" elven knowledge, which is not how the game worded it. But the point still stands even if the Veil Jumpers, for some condescending plot reason, completely lost control of this knowledge, or were so flippant as to put everything on Thedas' wikipedia without curing it at all.

Let's accept, too, that the Archive contains knowledge of how to build something equivalent to nuclear weapons, which one could argue is in fact truly dangerous, but... Well. Do you think it's fair that the countries that have nuclear arsenals are some of the most vocal about the dangers of other countries ever developing their own?

Because that's what it feels like, to me, when the game calls elven knowledge dangerous without ever allowing you to question -- what about Tevinter rituals and magic? Tevinter's millennium of slavery, still in practice at present day? Should we destroy all their libraries too to keep the world safe from dangerous magics? Why do we only get to tell the Dalish, the nomad nations severely subjugated in present Thedas (If you ever played the previous games and have the context, at least, since this game that happens in Tevinter somehow manages to completely gloss over racism against elves as if it never existed) to destroy a one-of-a-kind, ancient trove of knowledge? And have it be framed as good and safe? As "moving forward"?

If you choose to free the archive, Rook says "The elves deserve the chance to chart their own course" to which Bellara answers "Right. Define ourselves by who we are, not who we were," but once again that writing just makes me question Bioware -- Do they not understand the point of history at all? Do they think indigenous peoples are monoliths stuck in the past if they choose to study the history they lost to colonialism? What purpose do they think that keeping that history and culture extinct serves? Who do they think it benefits?

If you step outside of what the game is telling you as fact and think for yourself, with the context of the other DA games in mind, do you still agree that it's inherently dangerous to keep the Archive? Do you still think these are equally morally weighted choices?

Or would you agree that DAV has to subtly convince you, out of character, that keeping this knowledge is inherently dangerous to make this dichotomy make sense?

Again. This wouldn't bug me if they just owned up to the fact your protagonist can, once again, genocide elves/their culture, just like in previous games. And scapegoat present elves too for the sins of their thousand-years-old tyrants, now suddenly returned (it would make so much sense for characters in the narrative to scapegoat the elves, and for us as heroes to fight against that. But no, they don't even go there except through Bellara's guilt.). It's just bizarre to have an elven historian guiltily agreeing with destroying the Archive and then telling us "The Evanuris broke us and kept us broken" without anyone, either Rook or her, ever mentioning a thousand years of Tevinter slavery and several centuries of Andrastian persecution and subjugation.

No. The Evanuris are the be-all and end-all of evil and everything bad that ever happened in Thedas, ever, can be traced back to them.

[elbiotipo]

I would also say "don't send people death threats about AI" but I think many people are literally not ready for this conversation

[niobiumao3]

Another way TBB S3 was disappointing to me: Crosshair's redemption arc goes into how loyalty is one of his big motivators, yet Tech's loyalty to him when he insisted they go after Crosshair and subsequently was lost on that same mission is NOT ONCE BROUGHT UP.

The number of ways the emotional impact of Tech's death is just flat out ignored is really grating to me and it's why I struggle to enjoy the season at all. Not even just his absence--though ofc that sucks too--but that his absence is only felt as a series of missing skills and in random asides. Where's Crosshair feeling especially raw over the fact that the moment Tech knew Cross was still loyal to them he was ready to come and fetch him, and died trying?

Never not going to be bitter about all the emotional beats which were thoroughly and completely ignored.

[tgirlschlatt]

clingy bdays being literally right next to each became so true to me that i literally forgor tubbo actually shares a join day w/ fundy. kinda funny. you and your fellow cabinet traitor bestie. you and your fellow 'overlooked for tommyinnit' complex haver that responds to it in completely opposite ways to you. dreamon hunters. single parenting in the snow. shit happens and you think cruel things about each other sometimes but you're still cool ultimately. interesting!

[essektheylyss]

Brennan describes the inside (for lack of a better word) of the Occulus Thalamus as "a swirling nebula of glimmering potential, cresting eternally beyond the threshold of being," before turning instead to describe, for ease of mortal imagination, a physical palace.

I am curious to know if this, like other aspects of Aeorian technology that we've seen, is some kind of dunamantic distillation; if the "infinity" that is described, this monumental repository of data, is in fact an innovation created from Aeor's technological explorations of a beacon and dunamis itself. It would certainly negate any of my idle considerations of how you'd even store that much data, given the space(s) inside the beacons are suggested to be fairly infinite.

It would also track as a method of collection, given the other uses of the beacons in Exandria as a method of holding a soul and then returning it to another vessel, as well as its use for observation and learning; the Dynasty uses consecution for gaining perspective and learning about the world, and the Aevilux apparently use their own rituals as a method of external observation of themselves for learning who they are. Both of these have their downsides, of course, but overall they seem to be employed for thoughtful introspection.

If the Occulus Thalamus was based on the same concept and arcane background, it certainly showcases the dark side of the impulse to observe and learn and record, in excess and without consent. The same drive to learn about the world, to understand those outside of oneself, to connect with the world, used instead for all-encompassing surveillance and control.

[goodthingscomeinthrees]

not to sound all "old man yells at cloud" but this nymag article is genuinely frightening. it's like no one wants to think anymore!! if you can't do anything without consulting some kind of ai, you're going to atrophy your critical thinking and problem-solving skills entirely.

i'm so incredibly glad i finished uni in 2020, before *waves hands* any of this. i've never used an ai bot, ever, and i don't plan on changing that anytime soon.

[the-bi-space-ace]

It has been a little bit since tbb ended and even longer since “Plan 99” aired and I’m still sitting here completely crushed. I miss Tech. I miss him and I’m still so upset and I don’t want to be a complainer but I’m still so gutted by season 3 and *gestures to everything* and I just feel so emotional about it.

[laughhardrunfastbekindsblog]

You know the one point above all others that leaves me convinced Bad Batch season 3 underwent some massive rewrites?

The time skip in episode 1.

Now, the time skip could have worked fine if there had been any attempt later in the season to meaningfully follow up on the vital conversations that apparently took place during said time skip. But that's not what happened.

And I have too much confidence in the writers' vision/abilities to believe they originally intended to use the time skip the way it ended up being used: to completely gloss over the aftermath of Tech's fall to the point that the audience is left wondering - maybe Tech's family/friends have already processed it and moved on? Or maybe they haven't?? Who knows! Let's leave it super vague all season long and have the audience interpret it as they will! Pick your preferred grieving method and tell yourself that's what all these characters did during the time lapse, or if that doesn't work for you then just "something something stoic soldiers."

To give a clear example of what the writers are capable of: Mayday has the distinction of being recognized as THE tipping point to Crosshair finally turning on the Empire and later is given a satisfying, if heartwrenching, callback scene that decisively provides closure for his loss. Remember, Mayday is a character in ONE episode. Just ONE. In the grand scheme of the show, he probably qualifies as a tertiary character. Crosshair knows him for, what, 2 days at most? And yet Mayday is still definitively recognized as a key influential figure in Crosshair's life.

I love Mayday. He deserves all the recognition and more. I bring all this up simply to compare to how the show handles Tech's death, especially for Crosshair.

Tech is Crosshair's brother, was raised with him from birth and lived and worked with him day in and day out for over a decade, and for years they were in life-and-death situations together. Unlike with Mayday, Crosshair wasn't there when Tech died - died on a mission he had pushed for to save Crosshair from consequences of his own choices. Not only was Crosshair not there for Tech in his final moments, but the last time he saw Tech, Crosshair was arguing with him along with the rest of his brothers. Vitally important as Mayday is to Crosshair, Tech is even more so (or should be). Given all this, I'm supposed to believe the writers' grand plan all along was to skip over the critical moment where Crosshair finds out about Tech, spend the rest of the season ignoring all other opportunities to address it, and throw in one line during the finale ("Clone Force 99 died with Tech") that somehow manages to simultaneously deprive us of any semblance of catharsis for Crosshair AND completely miss the point of why Tech had sacrificed himself in the first place??

Nope. I don't believe it. There were forced rewrites on a time crunch. I REFUSE to believe the writers responsible for the near-perfection that is Bad Batch seasons 1-2 would, on their own, so thoroughly botch something as crucial to the show as Crosshair dealing with Tech's (supposed) death. There had to have been some kind of outside interference.

(I am clinging to the theory that the rewrites were part of a bigger plan to save some plot points for continuation in another project; but the point still stands that there had to have been significant rewrites in the first place.)

And since there would have been little to no reason to take out scenes with proper closure for Tech's fall during the rewrites if the original intention was indeed for Tech to be dead, I conclude yet again that Tech isn't actually dead.

I will say this for the time skip: it is what first pushed me into writing Bad Batch fanfiction. So there's that.