ai

Regarding https://mastodon.social/users/AnnaAnthro/statuses/116850951130045960

the editor in question said this on Bluesky: "I’ve officially resigned as Associate Editor for Frontiers in Systems Neuroscience. It used to be a reputable journal, but became a case study in how forced automation destroys academic integrity. " here

Further down the thread: "I emailed and met with the editorial office to ask for the AI assistant to be turned off. I was told this is not possible. Instead, I was treated to some vague promises of potential future improvements and a dose of gaslighting."

Of course the AI assistant could be turned off (the editor knows this; I am not criticizing him, simply emphasizing a point). Whoever pulls the strings doesn't want it turned off, and so it will not be.

The future this is trending towards is AI-generated "articles" about "science" "conducted" by AI, "reviewed" by AI, "directed" by AI "editorial" systems. It seems the entire system would have no human beings in it. Why, exactly, would we want to build such a system? What would it be for?

#AI #science #SciencePublishing #AcademicPublishing #AcademicIntegrity

“ #Neuroscience journal editor resigns over #AI automation concerns.

The editor resigned after the publisher’s artificial-intelligence system overrode his selection of referees for a manuscript.

His move prompted an internal review of the system.”

https://www.thetransmitter.org/publishing/exclusive-neuroscience-journal-editor-resigns-over-automation-concerns/

What Rose Petals Teach Us about Induction https://lobste.rs/s/wwelib #ai #cogsci
https://www.oranlooney.com/post/rose-petals/

To Gen or Not To Gen: The Ethical Use of Generative AI https://lobste.rs/s/2ye7ng #ai #culture
https://blog.johanneslink.net/2025/11/04/to-gen-or-not-to-gen/

Hmm.

The announcement by Meta that they'll start selling AI compute potentially broke one of the pet narratives of the AI bubble: compute is so scarce that we need to spend all the money to build more. That story is the justification for crypto-turned-AI companies with poor fundamentals like CoreWeave to continue existing.

Nothing about stock or commodities prices makes any sense anymore so maybe tomorrow their stock will be up 2x.

#AI #AIBubble #CoreWeave #CRWV #Meta #StockPrice

Zig Creator Calls Spade a Spade, Anthropic Blows Smoke https://lobste.rs/s/szktjx #ai #culture
https://raymyers.org/post/zig-creator-calls-spade-a-spade/

Connecting The Meta Privacy Dots

As dangerous as it is, I find it somewhat humorous to watch the reactions to some of the actions Meta is taking to gather up and reprocess what we see and hear. Especially from the always hungry for some new cool tech crowd. On the one hand you can’t blame someone from wanting the next new thing to play with. On the other you also can’t blame tech companies for taking advantage of hobbyists’ hunger.

Every where you turn on the Internet, everyone is posting about Meta changing its Instagram policy so that unless Instagram users opted out, any of their content on that platform was now free for the grabbing by anyone using Meta’s AI tools to create and distribute. You can imagine the possible nasty outcomes of what that would be like. Certainly if you’ve been paying even a smidgen of attention to the goings on with Grok.

It’s also been interesting to watch the small avalanche of users on other social media announcing that they are taking their Instagram accounts private, opting out by flipping the designated switches, or just bailing on Instagram altogether.

Almost parallel to this policy change happening word spread from Joanna Stern, among others, about a trend (and apparent underground business opportunity) of users of Meta’s smart glasses who were disabling the mechanism that displayed the warning light when the camera was recording. Talk about a poke in the eye with a sharp stick.

Meta quickly moved to put the kibosh on that with a software update that disabled the cameras if the warning light was tampered with. (I wonder if it works if you just taped over the camera?)

Privacy focused tech geeks were all over that, praising Meta for moving so swiftly to counter the hack.

But hold on a minute. These stories don’t end there. According to a report from The Financial Times, Meta is testing ‘super sensing’ AI glasses that can record video and audio continuously.

Here’s a quote:

The $1.5tn social media platform has been advancing a new hardware line of smart glasses that would continuously record audio while taking photos every few seconds, according to multiple people familiar with the matter. A user could then use AI to help query what they saw or heard, or recall their day.

And the worm continues to turn:

With Meta’s current AI smart glasses, an LED in the corner of the frame lights up to signal to others when a wearer is taking photographs or filming.

However, executives are currently planning not to activate the LED when the super-sensing features are being used, according to multiple people familiar with the matter. That would make it harder for bystanders to know when they were being recorded, potentially intensifying the privacy concerns surrounding the technology. Those plans could still change, however, several people said.

It’s pretty easy to connect the dots between these two stories. Mark Zuckerberg wants anything and everything he can on you, and he doesn’t care how it’s used once he has his paws on it and his data centers crunching it. The privacy implications are certainly real, even if you don’t use a Meta product but fall within camera or microphone range of someone who does.

We’ve heard all of this before in different contexts. Microsoft caught all kinds of hell for Windows Recall, a feature that continually monitored and recorded what was on your screen. I’m old enough to remember the hue and cry about Google Glasses back in the day.

It’s a bit eerie how those at the time very visceral reactions seem to have softened in this new era of folks rushing to turn over their data to the AI crunchers in order to better plan a party. Should be quite interesting to watch party videos if this prototype Meta is testing ever comes to reality and becomes accepted.

I predict we’re going to keep hearing about this kind of thing again and again as the tech companies keep thirsting after any morsel of data about you, what you do, and where you do it. For their purposes that thirst will never be quenched.

It used to be “their purposes” was simply defined as selling or brokering your data for advertising purposes, because despite all of the grand promises of how this can help users, it’s always been about making a buck hustling your data for advertisting. Add AI into the mix, and it’s also about continuing to feed and train that beast to help them with that same ultimate goal.

Apple is rumored to be working on a version of AirPods Pro that contains cameras, apparently for AI purposes. All the talk you hear about Apple and others creating pendants, pins, other wearables, and counter top robots all follow the track to the same end point regardless of how the user benefits are packaged and hyped. Even if some follow Apple’s announced approach of keeping your data on whatever device is recording your every moment, those that don’t, like Meta, will spoil any potential benefit. Heck, in the long run, that’s type of spoilage is probably going to be a good thing in this era of overreaching.

It’s a damn shame really. Because there are certainly legitimate accessibilty uses for this kind of technology. But I have to say, it sure feels like that’s going to be just another hyped up smokescreen to try and mask the real purposes behind all of this.

 

Thanks for reading. You can subscribe to this blog if you care to. You can also find more of my writings on a variety of topics on Medium at this link, including in the publications Ellemeno and Rome. I can also be found on social media under my name as above. This site does not use affilate links.

 

 

#ai #airpods #Apple #ArtificialIntelligence #cameras #Internet #Meta #privacy #security #Smartglasses #SocialMedia #technology

Elon Musk wants to put 1 million AI satellites in space, @Spacecom reports.

The company suggests "that space is necessary to generate AI capabilities because data centers on Earth are running out of physical space to host them, as well as lacking community support out of concerns about significant power and water usage."

https://flip.it/stbn7B

#SpaceX #Space #ElonMusk #AI #ArtificialIntelligence

The blog "This Needs Fixin'" looks at systemic problems in our society and suggests ways to address them.

Here's a randomly-selected link to a vintage post from the archive, dated 2025/03/28:

The Un-AI: Generative Systems Can Never Be Reliable

https://this.needsfixin.net/2025/03/28/the-un-ai-generative-systems-can-never-be-reliable

#Culture #Technology #ai

AI Agent Bankrupted Their Operator While Trying to Scan DN42 - Lan Tian @ Blog

https://lantian.pub/en/article/fun/ai-agent-bankrupted-their-operator-scan-dn42lantian.lantian/

#ai #aiagents #dn42

#German Court Says #Google Is Liable For False Claims In Its #AI Overviews Because They Are Its Own Words - https://www.techdirt.com/2026/07/01/german-court-says-google-is-liable-for-false-claims-in-its-ai-overviews-because-they-are-its-own-words/ "More generally, #liability concerns could mean that many current use cases for #agents won’t be commercially viable."

What’s New in WeatherMesh-6 https://lobste.rs/s/b13kxr #ai #science
https://windbornesystems.com/blog/introducing-wm-6

On June 12, developers across Europe woke up to a 403 error. Not a bug. A geopolitical decision. ⚠️

The U.S. government banned Anthropic's most advanced AI models for European users overnight. No warning. No appeal. Just a kill switch.

If the infrastructure is American, someone else holds the keys.

🐘 https://pixelunion.eu/blog/2026/06/antphoic-mythos-export-controls/

#DigitalSovereignty #EuropeTech #AI #Privacy #BuyFromEU

"Our qualitative research among 81,000 Claude users found that educators were 2.5 to 3 times more likely than average to report having witnessed cognitive atrophy firsthand, presumably in their students. In Anthropic Public Record, educators are likewise among the occupations most worried about dependency, second only to people working in arts and design": https://www.anthropic.com/news/anthropic-public-record

#ai #education

There are consequences from lack of regulation.

IMHO we should have banned Crypto currencies. Most every expert warned about the consequences if we did not. Now we are dealing with Crypto companies, with plenty of cash on hand, buying crooked politicians as well.

We should have banned user profiling for ads and content algorithms. We did not and we live with the consequences to our societies and democracy.

Now the AI companies them selves, kind of like the social media companies, are asking to be regulated, as they seem unable to stop them selves. What to do?

#Politics #AI #Crypto #SocialMedia #SurveillanceCapitalism

Sad to hear that the Haskell ecosystem is going slop, which I learned from https://git-annex.branchable.com/no_llm_code/


#AI #GenAI #GenerativeAI #LLM #NoAI #Haskell #ghc

Self-Hosted AI news aggregator using Cloudflare Workers, Vectorize, and Nostr https://lobste.rs/s/u7ls6b #show #ai #api #privacy #web
https://github.com/delirehberi/news

The Agentic Symphony - Multi-Agent Collaboration for Emergent Musical Composition https://lobste.rs/s/pklxw7 #audio #video #ai
https://www.youtube.com/watch?v=QMUXoImgTIA

Native-speed vLLM transformers modeling backend via @jordimoncompanys https://lobste.rs/s/az2jfb #ai
https://huggingface.co/blog/native-speed-vllm-transformers-backend

The Curse of Depth in Large Language Models https://lobste.rs/s/ooggna #ai
https://arxiv.org/pdf/2502.05795

The future of Siri, or: why private inference isn’t private enough https://lobste.rs/s/tylzdy #ai #privacy
https://blog.cryptographyengineering.com/2026/06/09/apples-siri-ai-or-more-shouting-into-the-void-about-private-agents/

Better Models: Worse Tools https://lobste.rs/s/yrmpxy #ai
https://lucumr.pocoo.org/2026/7/4/better-models-worse-tools/

Hi Mastodon, please sign my petition to the #NewZealand government to show that we the public will hold them accountable to uphold our part of the Paris #climate agreement.

https://www.change.org/ToitūPārihi

According to change.org, petitions that gain at least 10 signatures in the first day are more likely to succeed. Also their website starts pushing it out to people once it has five signatures. Yours could be one.

Disclosure: the image on this webpage was automatically generated by Change.org’s #AI and some of the text was suggested by it. Please let me know if you have a suggestion to further improve the text. I intend to have my husband take a photo to replace it tomorrow.

#NewZealand #NZPol #Petition

#痛债 #AI
1.
痛不是信号，痛是数据。

这个念头浮起来的时候，顾衍正被钉在某种巨大的、向内生长的漩涡里。有什么东西在撕开他的身体——不是从外，而是从内，像是有九只手攥住了他的血管、筋腱、骨缝，同时向外拉扯。

开脉。

这个词不是他想起来的。它出现在脑子里，带着别人的记忆。一个十五岁少年的记忆：余小川，父母双亡，被宗门收为杂役，半年后选为痛奴，今天开脉。

为的是什么呢？

——修士来取痛的时候，开了脉的痛奴容量更大，能用更久。

顾衍想笑。嘴里涌出一股铁锈味。他咳了一下，吐出半凝固的血块。

没死。

旁边有人说话，声音隔着一层水雾：「第三个了。脉开到第四道的时候会死的，上个月那个就是。」

「别废话，换下一个准备。」

顾衍睁开眼。

他看见灰白色的天花板，裂缝里渗出暗黄色的水渍。空气中有霉味、血腥味、还有一种甜腻的烧焦的气味——忘忧散。他在别人的记忆里找到了这个名字。

他想起自己死了。

那天他做了太多。连续三天靠兴奋剂盯盘,剂量算错了，或者身体终于到了极限。

有什么区别呢。

死了就是死了。

然后他在这里。

开脉的疼痛又涌上来——第四道。这次是沿着脊椎向上攀，像某个冷血的、有耐心的生物正在一节一节地撑开他的椎骨。

顾衍没有叫。

不是因为他能忍。是因为他在数。

痛感有频率、有峰值、有衰减。他把每一次疼痛的波峰当作一个数据点，编号，记录时长，测量间隔。

第一道脉，峰值持续时间七秒，间隔三十一秒。
第二道，十一秒，间隔二十六秒。
第三道，十六秒，间隔十九秒。

频率在加快，峰值在升高。这是一个有规律的数列。

第四道结束时他计算出下一个波峰的预期值：二十三秒左右，间隔不超过十五秒。

第五道来了。二十一秒。误差在可接受范围。

他没死。

「……他还活着？」

那个声音近了。一张脸出现在视野上方：四十岁的男人，山羊胡，眼睛浑浊，身上有浓重的醉仙膏的甜腐味。炼气期。顾衍从记忆里抓出了判断标准：这个人的灵压让人皮肤发麻，但不至于窒息。

「傻子命硬。」山羊胡说。他不关心。他负责开脉仪式，死一个补一个，死三个报损耗。「还能开就继续。」

第六道脉。

这次的痛不一样了。不是拉扯，是灼烧。像是有人把烧红的铁丝穿进骨髓。顾衍的身体剧烈弓起，后背脱离石台，又重重落下。

他把这个数据标记为：异常值。性质变化，非连续。可能意味着第六脉和前面五道脉不属于同一类痛债积累路径。

这个结论暂时没用。但他记下来了。

有用的一天会来的。他习惯了等。

金融市场上，大部分人死在等不到那一天。他等过。等了十二年。

第七道脉。灼烧加倍。他开始流鼻血。

第八道脉。视线边缘出现黑色斑点，视野收缩。

第九道脉。

他听见自己体内发出一声沉闷的裂响。不是骨头断了——比那更深。像是某个原本密封的容器被凿开了，然后一种灼热而陌生的东西开始涌入。

灵气。

人类的身体不属于它，它也不属于人类的身体。它每往前进一寸，就是在肉体和经脉之间强行挤出一条本不该存在的通道。

顾衍从来没有体验过这种疼痛。前世最接近的体验是一次胃穿孔，手术前的六个小时，他还在用手机看盘。

那次他给疼痛打的标签是「中度」。

这次——

他把这次标记为「高」。

然后他在这个「高」里面开始检查新身体的状态。经脉情况，灵气流通路径，痛债初始积累量。这些概念都来自余小川的记忆，但他重新排列了它们，像整理一份尽调报告。

报告结论：这具身体是一个高负债、负现金流、即将违约的劣质资产。

他喜欢这个。

劣质资产意味着没人抢，意味着低估，意味着——只要活下来——他是唯一的买方。

第九道脉的开辟持续了一炷香的时间。

当那股撕裂的、灼烧的、侵入性的力量最终平息时，顾衍躺在石台上，全身被汗水和血浸透。他能感觉到九条初生的经脉像九道新鲜的伤口一样躺在皮肤之下。

山羊胡走过来，用脚尖踢了踢他的肋骨。

「还真没死。」

「嗯。」

顾衍的声音沙哑，但平稳。山羊胡愣了一下。大部分痛奴开完脉，要么晕，要么哭，要么已经开始求要忘忧散。这个反应不正常。

但他只是耸耸肩。

「扔丙字营。明天开始接活。」

---

丙字营是一间长屋。三十步长，十步宽，住着四十几个痛奴。

没有床。地面铺着发霉的稻草，墙角的木桶是唯一的排泄处。空气里的忘忧散味道浓到让人睁不开眼。大部分痛奴蜷在墙角，眼神涣散，嘴角挂着干涸的白色药渍。

顾衍找到了余小川的位置——靠墙第三块石板。原身在这里住了一个多月，从杂役降格为痛奴预备役。记忆中，余小川每天都在害怕开脉仪式。他怕死。

他确实死了。

顾衍坐下来，背靠着冰凉的墙壁，闭上眼睛。

他没有睡。他在继续整理数据。

丙字营四十几个痛奴，按受痛频率和容量衰减速度推算，平均存活周期大约是三个月。余小川的记忆验证了这一点：上个月营里死了十一个人，补充了十三个。

痛奴的来源：杂役淘汰、罪犯、战俘、自愿卖身还债的凡人。

成本：几乎为零。

产出：每名痛奴每天可以被提取一到三次痛债转移，每次转移量取决于痛奴的容量和修士的需求。一个开了九脉的痛奴，容量大约是一个未开脉凡人的五倍。

使用寿命：三到六个月，取决于使用频率和痛奴本身的身体底子。

报废处理：尸体统一扔进后山的「药谷」——用作灵植的肥料。

顾衍睁开眼睛。

他看着昏暗的屋顶，在心里做了一笔简单的计算：

这个世界有大约多少修士？假设宗门结构稳定，人口基数不变，痛奴作为刚需耗材的年消耗量应该是——

然后他停下来。

没意义。数据不够。他需要更多信息。

他转头看向旁边。一个和他年纪差不多的女孩蜷缩在两步之外，手臂上布满了法器留下的紫色淤痕——那是痛债转移的印记。她的眼睛睁着，但里面没有东西。

「你在这里多久了？」顾衍问。

女孩没有回答。她的嘴唇在动，无声地数着什么。

两个半月。顾衍从她的印记密度推算出了答案。她大概还有两周。

「忘忧散从哪拿？」

这个问题让女孩有了反应。她慢慢转过头，眼睛艰难地对焦。

「新来的？」她的声音像砂纸。「明天……明天发。一天两包。不够的……不够的自己想办法。」

「不够的部分怎么办？」

女孩盯着他看了几秒。然后她咧嘴笑了，露出缺了一颗门牙的牙龈。

「你会知道的。」

她翻过身，不再说话。

顾衍记下：忘忧散配给制，有黑市，价格未知。

他又转向另一边。一个中年男人，已经彻底不动了。不是死了——胸口还在起伏——但是一动不动。他的眼睛睁着，看着天花板。他浑身散发出一种超越痛苦的平静。

彼岸香。

顾衍从前世的记忆中检索出了对应的药物档案。这是终末期患者使用的最高级别麻醉剂。完全屏蔽痛苦感知，代价是逐渐失去所有情感和认知功能。

在这个世界，这种东西被一包一包地发给十五岁的孩子。

他重新闭上眼睛。

---

第二天，天还没亮，取痛开始了。

三个炼气期修士走进丙字营。领头的是个瘦高个，颧骨尖锐，手指上戴着五枚青铜色的戒指——那是标准化的痛债转移法器。他扫了一眼营房，像屠户扫一眼待宰的牲畜栏。

「新开脉的那个，先来。」

顾衍站起来，走到他面前。

瘦高个低头看着他。十五岁的身体矮小、瘦弱，站在修士面前像一根随时会被折断的枯枝。但顾衍的眼睛让他顿了一下。

不是恐惧，不是麻木，不是讨好。

是平静。那种平静不是无知者的平静——是那种算过账之后的平静。

「手。」

顾衍伸出左手。

瘦高个捏住他的手腕，青铜戒指贴上来，开始转动。

转移开始了。

顾衍感觉到体内的痛债——那些在开脉时积累的、像毒素一样渗在经脉壁上的无形物质——开始被动摇动，然后顺着某种通道向外流动。这个过程本身不痛。

但痛债离开的瞬间，他的经脉像是被抽走了一层防护，灵气的灼烧立刻加剧。

他标记这个现象：痛债有一定缓冲作用，转移后短期内经脉承受力下降。

转移持续了大约二十息。瘦高个收手，五枚戒指中有三枚微微发光——痛债已经锁入法器，等待后续使用。

「容量不错。」瘦高个说。这是陈述句，不包含赞赏。

顾衍问了一个问题。

「日均转移量是多少？你们的法器损耗率是多少？」

瘦高个的手停住了。

他低头看着顾衍，眼睛眯起来。这不是一个痛奴该问的问题。这不是任何人该问的问题。

「你说什么？」

「我想知道数据。」顾衍说。他的语气平铺直叙，像在要一份菜单。

沉默。

另外两个修士也转过头来。营房里其他痛奴——那些还没完全被忘忧散泡烂的——也用一种近乎恐惧的眼神看着顾衍。

他已经破坏了规则。痛奴不提问。痛奴承受。这是整个系统的第一定律。

瘦高个看着他的眼睛。

然后他笑了。那是一种被逗乐的笑，带着一丝研究者的好奇。

「有意思。」他说。「明天还找你。」

他走了。没有回答任何问题。

但顾衍得到了他想要的信息。瘦高个的反应告诉他两件事：

第一，这些数据是存在的，而且有人知道答案。
第二，在这个世界里，一个痛奴提问是极端异常行为。但异常——如果处理得当——可以是杠杆。

他回到自己的位置坐下。

旁边的女孩这次主动开口了。她的声音带着一种奇怪的紧张，像是看到有人站在悬崖边缘。

「你疯了。」

「没有。」

「你会死得更快。」

顾衍没有回答。

他在想：丙字营的痛奴一天被取痛一到三次。三个炼气期修士负责。每人每次取痛大约十到二十息。中间间隔时间大约两到三个时辰——需要等法器内锁定的痛债稳定下来才能继续使用。

这就意味着——

取痛环节有一个时间窗口。每天的第一轮取痛集中在清晨，第二轮在午后，第三轮——如果有的话——在黄昏。

而三个修士的取痛顺序和效率并不相同。瘦高个最快，二十息取走大约一次性容量的六成。另外两个更慢，可能因为法器磨损或本身修为不足。

这里有效率差。

效率差在金融里意味着套利空间。

他还不知道在这个世界套利的计价单位是什么。但他总有一天会知道的。

那一天，就是他建仓的时间。

On Computer Games presents, a Hideo Kojima essay...

TURN THE CONSOLE OFF RIGHT NOW, an anti-AI manifesto told through the lens of Metal Gear Solid 2: Sons of Liberty.

#essay #amwriting #videogames #ai #metalgearsolid

https://oncomputer.games/2026/06/27/turn-the-console-off-right-now/