Alice Schwarze
@aliceschwarze.bsky.social
Head of Research @ Utah AI Policy Office // math PhD // networks, complex systems, machine learning, and all things AI // mom & cat lady
Reposted by Alice Schwarze
Marino Badiale, Isabella Cravero
Computations for the first Lyapunov coefficient
https://arxiv.org/abs/2511.08428
Computations for the first Lyapunov coefficient
https://arxiv.org/abs/2511.08428
November 12, 2025 at 6:34 AM
Marino Badiale, Isabella Cravero
Computations for the first Lyapunov coefficient
https://arxiv.org/abs/2511.08428
Computations for the first Lyapunov coefficient
https://arxiv.org/abs/2511.08428
Reposted by Alice Schwarze
November 12, 2025 at 6:35 AM
Reposted by Alice Schwarze
Panpan Chen, Nader Motee, Qiyu Sun
Beyond Carleman Linearization of Nonlinear Dynamical System: Insights from a Case Study
https://arxiv.org/abs/2511.07805
Beyond Carleman Linearization of Nonlinear Dynamical System: Insights from a Case Study
https://arxiv.org/abs/2511.07805
November 12, 2025 at 6:36 AM
Panpan Chen, Nader Motee, Qiyu Sun
Beyond Carleman Linearization of Nonlinear Dynamical System: Insights from a Case Study
https://arxiv.org/abs/2511.07805
Beyond Carleman Linearization of Nonlinear Dynamical System: Insights from a Case Study
https://arxiv.org/abs/2511.07805
Reposted by Alice Schwarze
Rafael Da C. Pereira, T\'ulio Vales, C\'assio H. Vieira Morais
Spectral decomposition and entropy for open set-valued maps
https://arxiv.org/abs/2511.07642
Spectral decomposition and entropy for open set-valued maps
https://arxiv.org/abs/2511.07642
November 12, 2025 at 6:36 AM
Rafael Da C. Pereira, T\'ulio Vales, C\'assio H. Vieira Morais
Spectral decomposition and entropy for open set-valued maps
https://arxiv.org/abs/2511.07642
Spectral decomposition and entropy for open set-valued maps
https://arxiv.org/abs/2511.07642
Reposted by Alice Schwarze
What Would it Take to Convince a Neuroscientist That an AI is Conscious?
I, @anilseth.bsky.social, and Michael Graziano weigh in:
gizmodo.com/what-would-i...
Thanks to Ellyn Lapointe for the opportunity to write about this.
I, @anilseth.bsky.social, and Michael Graziano weigh in:
gizmodo.com/what-would-i...
Thanks to Ellyn Lapointe for the opportunity to write about this.
What Would it Take to Convince a Neuroscientist That an AI is Conscious?
Before we can test for AI consciousness, we need to understand how consciousness actually emerges, experts say.
gizmodo.com
November 11, 2025 at 10:25 PM
What Would it Take to Convince a Neuroscientist That an AI is Conscious?
I, @anilseth.bsky.social, and Michael Graziano weigh in:
gizmodo.com/what-would-i...
Thanks to Ellyn Lapointe for the opportunity to write about this.
I, @anilseth.bsky.social, and Michael Graziano weigh in:
gizmodo.com/what-would-i...
Thanks to Ellyn Lapointe for the opportunity to write about this.
Luckily I am not in the business of fixing the US housing market but reposting because of the succinct policy analysis. Almost no policy is good for everyone or bad for everyone. Developing policy included identifying winners and loser and a hard honest introspection at how ones values each group
In an interesting turn of events, US and Danish government came up with the same to help with the affordability crisis: longer mortgages!
That has sparked some backlash. But in the US, it seems the discussion is focusing on the wrong stuff.
Well, let’s try to learn from 🇩🇰.
That has sparked some backlash. But in the US, it seems the discussion is focusing on the wrong stuff.
Well, let’s try to learn from 🇩🇰.
November 10, 2025 at 4:16 PM
Luckily I am not in the business of fixing the US housing market but reposting because of the succinct policy analysis. Almost no policy is good for everyone or bad for everyone. Developing policy included identifying winners and loser and a hard honest introspection at how ones values each group
Reposted by Alice Schwarze
This will spark some great papers: an open digital dataset of roads in the Roman Empire
www.nature.com/articles/d41...
www.nature.com/articles/d41...
‘Google Maps’ for Roman roads reveals vast extent of ancient network
A high-resolution digital map nearly doubles the known length of the ancient road network.
www.nature.com
November 9, 2025 at 1:38 PM
This will spark some great papers: an open digital dataset of roads in the Roman Empire
www.nature.com/articles/d41...
www.nature.com/articles/d41...
Terence Tao is more kind, patient, and generous than any of us deserves
Right, enough of James Watson - who's a senior academic you've met who's been an utter delight?
I'll go first: Jocelyn Bell Burnell
I'll go first: Jocelyn Bell Burnell
November 9, 2025 at 5:19 AM
Terence Tao is more kind, patient, and generous than any of us deserves
When I buy a trail mix with dried apricots, the apricots are delicious and everything else in it is just meh. When I buy a bag of dried apricots, the apricots are are moist and goopy not at all like the ones in the trail mix :-(
November 9, 2025 at 5:13 AM
When I buy a trail mix with dried apricots, the apricots are delicious and everything else in it is just meh. When I buy a bag of dried apricots, the apricots are are moist and goopy not at all like the ones in the trail mix :-(
Reposted by Alice Schwarze
I think if we went back to early structuralism (like Propp on the folktale) this is one of the insights that gets them started: the feeling we have that these stories are not governed by ordinary logic but by some weirdly rigid principle where things happen three times, hair must be mentioned, &c
November 9, 2025 at 2:52 AM
I think if we went back to early structuralism (like Propp on the folktale) this is one of the insights that gets them started: the feeling we have that these stories are not governed by ordinary logic but by some weirdly rigid principle where things happen three times, hair must be mentioned, &c
Reposted by Alice Schwarze
10% of their revenue. meta made $16 BILLION on fraud last year. and because these were a paid service that means that far more than $16 was lost by those defrauded
November 9, 2025 at 4:44 AM
10% of their revenue. meta made $16 BILLION on fraud last year. and because these were a paid service that means that far more than $16 was lost by those defrauded
Somehow believe that math, science, society would be so much more advanced if all humans were strictly incapable to count to or mentalize any number greater than 20.
November 5, 2025 at 5:08 PM
Somehow believe that math, science, society would be so much more advanced if all humans were strictly incapable to count to or mentalize any number greater than 20.
Me: I would like to save for retirement
My 2-yo: F that I am going on a 100% Parmesan diet
My 2-yo: F that I am going on a 100% Parmesan diet
November 5, 2025 at 3:39 PM
Me: I would like to save for retirement
My 2-yo: F that I am going on a 100% Parmesan diet
My 2-yo: F that I am going on a 100% Parmesan diet
Reposted by Alice Schwarze
Der @erc.europa.eu fördert Projekte, die besondere Innovationen versprechen. Die DIW-Forscher @hannesullrich.com und @pmlersch.bsky.social haben uns erzählt, wie ihre Forschung davon profitiert hat. Das komplette Interview gibt's auf unserer Jubiläumsseite: www.diw.de/de/diw_01.c....
November 4, 2025 at 1:39 PM
Der @erc.europa.eu fördert Projekte, die besondere Innovationen versprechen. Die DIW-Forscher @hannesullrich.com und @pmlersch.bsky.social haben uns erzählt, wie ihre Forschung davon profitiert hat. Das komplette Interview gibt's auf unserer Jubiläumsseite: www.diw.de/de/diw_01.c....
Reposted by Alice Schwarze
My new field guide to alternatives to standard LLMs:
Gated DeltaNet hybrids (Qwen3-Next, Kimi Linear), text diffusion, code world models, and small reasoning transformers.
🔗 magazine.sebastianraschka.com/p/beyond-sta...
Gated DeltaNet hybrids (Qwen3-Next, Kimi Linear), text diffusion, code world models, and small reasoning transformers.
🔗 magazine.sebastianraschka.com/p/beyond-sta...
November 4, 2025 at 2:49 PM
My new field guide to alternatives to standard LLMs:
Gated DeltaNet hybrids (Qwen3-Next, Kimi Linear), text diffusion, code world models, and small reasoning transformers.
🔗 magazine.sebastianraschka.com/p/beyond-sta...
Gated DeltaNet hybrids (Qwen3-Next, Kimi Linear), text diffusion, code world models, and small reasoning transformers.
🔗 magazine.sebastianraschka.com/p/beyond-sta...
Reposted by Alice Schwarze
Reposted by Alice Schwarze
This sounds cynical—but represents a huge advance over empty “AGI” speculation.
It’s a political question, not a technical one. Without social equality models *cannot do* many kinds of work (eg, negotiate agreements or manage workers). So they will only be human-equivalent if we decide they are.
It’s a political question, not a technical one. Without social equality models *cannot do* many kinds of work (eg, negotiate agreements or manage workers). So they will only be human-equivalent if we decide they are.
intelligence is the thing which i have. admitting things are intelligent means considering them morally and socially equal to me. i will never consider a computer morally or socially equal to me. therefore no computer program will ever be intelligent
November 2, 2025 at 11:01 AM
This sounds cynical—but represents a huge advance over empty “AGI” speculation.
It’s a political question, not a technical one. Without social equality models *cannot do* many kinds of work (eg, negotiate agreements or manage workers). So they will only be human-equivalent if we decide they are.
It’s a political question, not a technical one. Without social equality models *cannot do* many kinds of work (eg, negotiate agreements or manage workers). So they will only be human-equivalent if we decide they are.
Reposted by Alice Schwarze
Look folks gen AI videos are out in the wild so if you weren't already being careful about sharing rage bait videos then now is a great time to start
If you can't trace a video back to a trusted source then I recommend leaving it where it is, like a slice of bologna you found on the street
If you can't trace a video back to a trusted source then I recommend leaving it where it is, like a slice of bologna you found on the street
November 2, 2025 at 9:02 PM
Look folks gen AI videos are out in the wild so if you weren't already being careful about sharing rage bait videos then now is a great time to start
If you can't trace a video back to a trusted source then I recommend leaving it where it is, like a slice of bologna you found on the street
If you can't trace a video back to a trusted source then I recommend leaving it where it is, like a slice of bologna you found on the street
Reposted by Alice Schwarze
Reposted by Alice Schwarze
Adrienne Rich's poem is absolute genius.
About a decade ago there was already a quiz circulating (I saw it on twitter) of spotting real vs AI poems. I was 10/10 on it without trouble, but a lot of even Harvard educated coding friends were 5/10 chance. The end of art is when it can't be recognised.
About a decade ago there was already a quiz circulating (I saw it on twitter) of spotting real vs AI poems. I was 10/10 on it without trouble, but a lot of even Harvard educated coding friends were 5/10 chance. The end of art is when it can't be recognised.
A short essay following up on the recently rediscovered poem by Adrienne Rich, titled "Artificial Intelligence", and written in 1961. I do what she suggested back then. I forced AI systems to write some poems. rodneybrooks.com/a-prophetic-...
A Prophetic Poem about Artificial Intelligence Written in 1961 – Rodney Brooks
rodneybrooks.com
November 3, 2025 at 7:00 AM
Adrienne Rich's poem is absolute genius.
About a decade ago there was already a quiz circulating (I saw it on twitter) of spotting real vs AI poems. I was 10/10 on it without trouble, but a lot of even Harvard educated coding friends were 5/10 chance. The end of art is when it can't be recognised.
About a decade ago there was already a quiz circulating (I saw it on twitter) of spotting real vs AI poems. I was 10/10 on it without trouble, but a lot of even Harvard educated coding friends were 5/10 chance. The end of art is when it can't be recognised.
Reposted by Alice Schwarze
This paper is a great exposition of how "personhood" doesn't need to be, and in fact should not be, all-or-nothing or grounded in abstruse, ill-defined metaphysical properties. As I argued in my recent @theguardian.com essay, we can and should prepare now: www.theguardian.com/commentisfre...
[1/9] Excited to share our new paper "A Pragmatic View of AI Personhood" published today. We feel this topic is timely, and rapidly growing in importance as AI becomes agentic, as AI agents integrate further into the economy, and as more and more users encounter AI.
November 2, 2025 at 3:30 PM
This paper is a great exposition of how "personhood" doesn't need to be, and in fact should not be, all-or-nothing or grounded in abstruse, ill-defined metaphysical properties. As I argued in my recent @theguardian.com essay, we can and should prepare now: www.theguardian.com/commentisfre...
Checking references provided by Claude Opus. We are currently at 4:2 for non-existing vs existing papers. Not sure if I had gotten lucky with this in the past or if the 4.x models suck at referencing sources more than the 3.x models
November 2, 2025 at 3:45 PM
Checking references provided by Claude Opus. We are currently at 4:2 for non-existing vs existing papers. Not sure if I had gotten lucky with this in the past or if the 4.x models suck at referencing sources more than the 3.x models
I really need to go through the details of this paper because I trust math but also
November 2, 2025 at 3:12 AM
I really need to go through the details of this paper because I trust math but also
*sigh* looks like I need to spend more time monitoring robotics now too
Pay $500/month&request on phone app a task to be done-human virtually in your house teleops robot-sometimes succeeds. "In reality, projects like Tesla’s Optimus or 1X’s NEO are less about practical consumer robotics and more about cashing in off of technological hype." futurism.com/future-socie...
$20,000 Robot Servant Comes With a Major Catch
Owners of the NEO butler-bot from 1X will have to schedule a time for remote operators to plug in and complete tasks.
futurism.com
November 2, 2025 at 2:40 AM
*sigh* looks like I need to spend more time monitoring robotics now too
Now imagine the junior folks in labs whose careers depend on something working ...
Paul Nurse describing the main job of a PI
(From ‘The Thinking Game’, 2024)
(From ‘The Thinking Game’, 2024)
November 1, 2025 at 11:06 PM
Now imagine the junior folks in labs whose careers depend on something working ...