[Tsundoku]

Amazing what motion-capture can do these days.

[Azath Vitr (D'ivers]

Quote

recent research by scientist Tuhin Chakrabarty[...] has attempted to fine-tune large language models to produce better writing by feeding them authors' entire oeuvres. [...] when Chakrabarty ran the results by some creative writing graduate students, they preferred AI imitations of writers like Junot Diaz, Sigrid Nunez, and Tony Tulathimutte to the writers themselves, or could not tell the difference. [Critically acclaimed authors Vauhini Vara] and [Karan] Mahajan talk about their decades-long connection and familiarity with each other's writing. They muse on what it means that, when Vara talked Chakrabarty into letting her compete with a large language model, even Mahajan could not separate her original work from what it produced.

Literary Hub » Vauhini Vara and Karan Mahajan on When AI Tries to Sound Like Us

Here's a pre-print of the paper:

Readers Prefer Outputs of AI Trained on Copyrighted Books over Expert Human Writers

[Azath Vitr (D'ivers]

Quote

With help from AI, MIT researchers have discovered novel antibiotics that can combat two hard-to-treat infections

[...] drug-resistant Neisseria gonorrhoeae and multi-drug-resistant Staphylococcus aureus (MRSA).

[...] The top candidates they discovered are structurally distinct from any existing antibiotics, and they appear to work by novel mechanisms that disrupt bacterial cell membranes.

https://news.mit.edu...t-bacteria-0814

Generative AI may save your life. It's the most promising path forward (at least until we scale up quantum computing) for dealing with the surge in antimicrobial resistant infections:

Quote

One in six laboratory-confirmed bacterial infections causing common infections in people worldwide in 2023 were resistant to antibiotic treatments, according to a new World Health Organization (WHO) report launched today. Between 2018 and 2023, antibiotic resistance rose in over 40% of the pathogen-antibiotic combinations monitored, with an average annual increase of 5–15%.

[...] "Antimicrobial resistance is outpacing advances in modern medicine, threatening the health of families worldwide," [...] E. coli and K. pneumoniae are the leading drug-resistant Gram-negative bacteria found in bloodstream infections. These are among the most severe bacterial infections that often result in sepsis, organ failure, and death.

https://www.who.int/...otics-worldwide

And generative AI's ability to come up with new mathematical insights is now on the level of the world's best mathematicians. According to professor Ravi Vakil, Stanford University, president of the American Mathematical Society:

Quote

We integrated Gemini into various stages of the project to empirically test its effectiveness. In many instances, it proved useful in familiar ways: identifying connections to cross-disciplinary papers, writing data-generation code, and verifying minor lemmas. However, the most striking experience was how it propelled the project forward intellectually.

[...] As someone familiar with the literature, I found that Gemini's argument was no mere repackaging of existing proofs; it was the kind of insight I would have been proud to produce myself. While I might have eventually reached this conclusion on my own, I cannot say so with certainty.

https://x.com/A_G_I_...213692617285729

[Macros]

And hey, Grok in the Military can only go well

[Azath Vitr (D'ivers]

Macros, on 16 January 2026 - 11:15 PM, said:

And hey, Grok in the Military can only go well

Here's hoping the new Army of Shadows features incompetent bumbling hallucinating easily fooled Neo-Nazi robots and surveillance/PA systems praising Elon Musk as the Great Creator and Donald Trump as Jesus... they will be watching in your homes to make sure you're getting on your knees to pray to Donald Trump (as your Lord and Saviour), and not having any abortions in there.

But they'll probably add substantial guardrails to try to keep Grok from wreaking too much havoc:

Quote

Should the Pentagon proceed with the integration, the military will likely have to rely on additional guardrails and testing to prevent the same failures that have plagued the model's public deployments, sources told Information Security Media Group. Hegseth announced Monday that Grok will soon be integrated into military systems alongside other commercial AI tools as part of a broader push to accelerate AI adoption. Hegseth said the move is part of a wider "AI acceleration strategy" aiming to "unleash experimentation" [...]

Analysts said deploying Grok safely will likely require significant hardening, including sandboxed testing environments that mirror operational data, extensive red-team exercises designed to probe for failure modes and strict limits on what systems and datasets the model can access. Absent those controls, they warned Grok could introduce new attack surfaces into military networks, including exposure to prompt injection attacks, adversarial manipulation of outputs or unintended disclosure of sensitive context through model responses.

Pentagon's Use of Grok Raises AI Security Concerns

So those added vulnerabilities might be good news (for most of the rest of the world, if they can exploit them)...

This post has been edited by Azath Vitr (D'ivers: 16 January 2026 - 11:33 PM

[Azath Vitr (D'ivers]

Remember that "very soulful" hit song I posted two days ago?

Well, guess what...

Quote

A hit song has been excluded from Sweden's official chart after it emerged the "artist" behind it was an AI creation.

I Know, You're Not Mine – or Jag Vet, Du Är Inte Min in Swedish – by a singer called Jacub has been a streaming success in Sweden, topping the Spotify rankings. [...]

"The artist Jacub's voice and parts of the music are generated with the help of AI as a tool in our creative process."

Stellar said it was "first and foremost" a music company run by creative professionals, not a tech or AI outfit. The company added that creating a hit had required "something different" to simply prompting an AI tool to create a tune.

Partly AI-generated folk-pop hit barred from Sweden's official charts | AI (artificial intelligence) | The Guardian

Obviously calling it "I Know, You're Not Mine" could be interpreted as poking fun at the primary objection most people make to the most common forms of generative AI. Though since the two major generative AI music services, Suno and Udio, are both switching to "ethically trained" models (with consent of the rights holders) this year, and there are several other "ethically trained" music models for generating stems (individual instrumental or vocal tracks to incorporate as part of a song, intelligently responding to what's already been added to the song), that's soon no longer going to be much of an issue, so people who hate AI because it's AI, or because they're afraid of losing their jobs, will transition to other rationalizations to passionately glom onto.

From their statement (applying Google Translate to the Swedish):

Quote

"The work has been created through a focused, conscious and human-driven creative process, guided by a clear artistic vision. ...

We are strongly critical of the mass uploading of music, often referred to as "AI music slop," where anonymous actors or technology-driven entities upload thousands of songs without artistic intent. ...

[This song] is based on craftsmanship and contributions from experienced music professionals. The lyrics are inspired by events in our own lives and written from genuinely human experiences."

Jacub svarar om AI-låten som toppar svenska Spotifys topplistor: Vi är ett kollektiv bakom

Here's the song again for reference (yes, if you listen closely you can hear some artifacts in the voice):

[Tsundoku]

Here you go AVD, 2047. Set your alarm clock

AI bots make terrifying prediction of when machines will overtake humans as their ‘overlords’
AI bots have pinpointed the year they think they will overtake humans as our “overlords” in a terrifying prediction.

https://www.news.com...c230c369c70ab39

[Azath Vitr (D'ivers]

This seems appropriate for Valentine's day (video: magnetic modular robots connect with each other to transform into different configurations while performing different outdoor tasks):

https://www.facebook...e/v/1GgFjMtBfL/

(You don't have to sign in to watch, you can just close the window that asks you to sign in)

And maybe this one too ("Researchers built HumanX, letting humanoid robots learn from a single human video"):

https://www.facebook...e/r/1CMdZEKzU1/

Next they're going to teach them to f*ck...

Perhaps that's what I'd really like to see in the Olympics (winter or otherwise): humans vs humanoid robots vs non-humanoid robots. F*cking or doing whatever sports or feats of dexterity.

On the non-humanoid automation front, this seems like it might be a magical place for a date (and extra intimacy, without human servers to interrupt you... unless you like it when people watch):

Quote

A restaurant in China's Shenzhen is using magnetic levitation to serve food. Plates float along tracks to arrive smoothly in front of each diner.

https://www.facebook...e/r/1PxLR6SDvc/

[Azath Vitr (D'ivers]

The streets of Philadelphia... now have robots traversing them!

Quote

Customers can now choose a robot delivery instead of a human courier!
The electric, driverless robots developed by AV Ride are designed to help reduce emissions and ease road congestion

Spotted in Philadelphia - Video

Meanwhile in China---robots doing acrobatics, kung fu, moves that look like they're out of The Matrix (but it's robots in physical reality, not humans in virtual reality), and breakdancing:

Chinese robotics company Unitree just released a video of this large-scale demonstration of its humanoid systems - Video

This might be the full video from the Chinese Spring Festival:



... please hurry up and conquer liberate the United States already!

In all seriousness though, I was happy to read:

Quote

Robotics firms see backlog in orders after humanoids steal the show at Spring Festival Gala

Consumer interest in robots has surged in China since several leading models appeared on stage at Monday's Spring Festival Gala

https://www.scmp.com...g-festival-gala

But tragically people in the United States have been attacking and robbing robots:

Quote

Americans are unleashing their anger on food-delivery robots

The machines don't deserve it

A recent survey by Pew Research Centre found that Americans are far more concerned about AI intruding on daily life than people in other rich countries

[...] Videos of pedestrians assaulting them have gone viral. Some seem to be shaking them down for pad thai; others are releasing rage by tipping the poor things over.

Americans are unleashing their anger on food-delivery robots - The Economist

Clearly those delivery robots need righteous Chinese Kung Fu robots to protect them from evil American idiots trying to tip them over.

Meanwhile, non-robot agents have made major advances:

Quote

Here's the thing [almost] nobody outside of tech quite understands yet: the reason so many people in the industry are sounding the alarm right now is because this already happened to us. [...]

I am no longer needed for the actual technical work of my job. I describe what I want built, in plain English, and it just... appears. Not a rough draft I need to fix. The finished thing. I tell the AI what I want, walk away from my computer for four hours, and come back to find the work done. Done well, done better than I would have done it myself, with no corrections needed. A couple of months ago, I was going back and forth with the AI, guiding it, making edits. Now I just describe the outcome and leave.
[...]

The AI labs made a deliberate choice. They focused on making AI great at writing code first... because building AI requires a lot of code. If AI can write that code, it can help build the next version of itself. A smarter version, which writes better code, which builds an even smarter version. Making AI great at coding was the strategy that unlocks everything else.

Something Big Is Happening — matt shumer

This post has been edited by Azath Vitr (D'ivers: 18 February 2026 - 09:47 PM

[Azath Vitr (D'ivers]

[Grief]

Overall, the market looks hugely jittery right now -- but there's not even agreement about what to be jittery about. Is it overinvestment into AI, AI destroying the value of other investments, or somehow both at once?

AI firms got 61% of global venture capital funding last year. Meanwhile multiple investment banks are sounding the alarm about how exposed private credit firms are to software sectors that AI itself might disrupt.

You have what is essentially dystopian sci-fi moving markets, all while AI's impact on official data when it comes to say jobs or growth remains very limited.

When there is such huge uncertainty about the future it feels almost inevitable that the market will have it wrong.

[Azath Vitr (D'ivers]

Grief, on 26 February 2026 - 11:03 PM, said:

Overall, the market looks hugely jittery right now -- but there's not even agreement about what to be jittery about. Is it overinvestment into AI, AI destroying the value of other investments, or somehow both at once?

AI firms got 61% of global venture capital funding last year. Meanwhile multiple investment banks are sounding the alarm about how exposed private credit firms are to software sectors that AI itself might disrupt.

You have what is essentially dystopian sci-fi moving markets, all while AI's impact on official data when it comes to say jobs or growth remains very limited.

When there is such huge uncertainty about the future it feels almost inevitable that the market will have it wrong.

Axios on this today:

Quote

Substacks are the new short seller report in an environment where all the major banks are competing to land big AI IPOs.

Broadly, analysts have not been as forthcoming about any concerns they have on the AI trade.

Paul Kedrosky, a venture capitalist, tells Axios that's because banks want to remain in favor with AI firms so they can handle their IPOs and other deals.

Enter independent authors: their takedowns feel like clarity to investors who are steeped in uncertainty about AI's path forward. It's narrative roulette on Wall Street.

https://www.msn.com/...isk/ar-AA1X8ZGN

As far as the Saas-pocalypse goes, for companies with many years worth of secure proprietary data it seems extremely overblown---particularly companies that are likely going to be net AI beneficiaries like ServiceNow, Salesforce, and SAP.

Between the private credit shenanigins and Trump there does seem to be a heightened probability of a market crash this year... which is why I increased my weighting of CAOS and gold.

Meanwhile... some economists are finding AI's analytical capabilities extremely useful. John Cochrane (of Stanford's Hoover Institution):

Quote

I recently tried refine, an AI tool for refining academic articles ... I sent it the current draft of my booklet on inflation, to see what it can offer. I just used it once so far, with the free trial mode. I will be a regular user forever.

The results are stunning. The comments it offered were on the par of the best comments I've received on a paper in my entire academic career. And more concise and organized than the best ones. They aren't perfect, but the kind of analysis the program is able to do is past the point where technology looks like magic. I don't know how you get here from "predict the next word."

https://www.grumpy-e...st.com/p/refine