Lightnews — Scholar-powered news

Dan Sylveste

@constantinejn.bsky.social

110 followers 260 following 120 posts

Friend to monsters, companion to bees

Posts Replies Media Videos

Dan Sylveste

@constantinejn.bsky.social

⏺ These are substantial refactoring tasks. Let me spawn architect agents to analyze and plan

Stern faced main "from the future" style comic strip with a single tear and a pointer box saying SINGLE
TEAR
OF JOY

January 3, 2026 at 12:51 AM

Dan Sylveste

@constantinejn.bsky.social

The History Channel’s “Aliens” dude, in stylized black and white, meme pose, with the words "results were fudged a little bit" under him, in quotes

January 2, 2026 at 11:01 PM

Dan Sylveste

@constantinejn.bsky.social

Conclusion
I think Al agents are amazing as long as you are the driver & reviewing its work. Al agents become problematic once you re not making them adhere to engineering
principles that have been established for decades.
- S64 To gain insight into the current practice of agentic coding by experts, we studied experienced developers (3+ years of professional experience) through 13 field observations supplemented with a broader survey of 99 respondents. We aimed to learn about their values and motivations for using agents, workflow strategies, what agents are suitable for, and developer sentiments about using coding agents. We find that experienced developers do not currently vibe code. Instead they carefully control the agents through planning and active supervision because they care about software quality. Although they may not always read code to validate agentic output, they are careful not to lose control and do not let agents run completely autonomously, particularly having the agents only work on a few tasks at a time. They generally enjoy using agents, finding them suitable for accelerating straightforward, repetitive, and scaffolding tasks, including a large variety of software engineering tasks like writing tests, documentation, refactoring, and simple debugging and yet opinions are mixed about writing plans with agents, and they do not use agents for core business logic or complex tasks. AI capabilities and interfaces are changing fast: this work paints a picture of what is and is not working now, serving as a reference point both to calibrate expectations about current Al as well as to anchor comparison in future years to see how much Al has improved
since 2025. As of now, the Als are not taking over yet-experienced developers are still in control.

January 2, 2026 at 5:07 PM

Dan Sylveste

@constantinejn.bsky.social

January 2, 2026 at 2:03 AM

Dan Sylveste

@constantinejn.bsky.social

Behold

De Blasio meme from the Onion with “maybe you can” find a mayor who doesn’t suck shit.

January 1, 2026 at 10:54 PM

Dan Sylveste

@constantinejn.bsky.social

behold

a flow chart for finding the optimal architecture with a loop asking "good enough" as the exit condition

January 1, 2026 at 11:29 AM

Dan Sylveste

@constantinejn.bsky.social

January 1, 2026 at 11:12 AM

Dan Sylveste

@constantinejn.bsky.social

But this has downsides, too, and some of them are profound, and it's stupid not to acknowledge that. As Scott describes in Seeing Like a State, high-modernist agricultural schemes tried to simplify farming into clearly legible (abstracted) systems. The goal was to make for a more understandable, more coherent overall system that could be planned and controlled centrally, and eventually those system planners did learn (some things) and did incorporate (some of) the mětis that made local farming work. But it did starve people. And the people who literally starved to death did not exactly get to reap the rewards learned from their
loss. Nothing I'm saying should be taken to delegitimize the justified anxiety and anger of folks who are on the receiving end of the 90%-for-$20 implementation. Which should be mostly taken to mean the producers of work so disrupted, it's probably pretty likely to make the consumption of that work worse, too. But I'm describing, I'm not prescribing. Whether or not it will be net- good or not-and I am certainly plenty ambivalent about parts of it if not the whole-I do
believe that it will happen. And that's just the things that do work, isn't it? Like, the way companies are going to try to use these tools, fueled in no small part by frontier model providers' breathless hype, is going to hurt people, already has hurt people. I do think creative destruction is real and (again) on balance positive; I suspect that my team at work is going to accelerate substantially in 2026
after some pretty heavy layoffs late in 2025, and our LLM strategy is a lot of why I feel that way. But it's inhumane not to acknowledge that yeah, businesses might like the "was impacted"
language for a layoff but there's a moral obligation to regard it more concretely than that.

December 31, 2025 at 3:44 PM

Dan Sylveste

@constantinejn.bsky.social

December 31, 2025 at 1:41 AM

Dan Sylveste

@constantinejn.bsky.social

Baby, we are so back

Glenn Beck Posts Utterly Bizarre Sit-Down Interview With George
Washington Al Bot
Glenn Beck on Saturday revealed he's leveraged the power of artificial intelligence to interview the first President of the United States.
&www.yahoo.com

December 28, 2025 at 6:27 PM

Dan Sylveste

@constantinejn.bsky.social

Lost three tree across the street in that massive storm.

Sure opened up the view, though.

My word.

2 out of the three tree that were lost , coming down quick.

viewing a razor sharp horizon of the pacific now that the view has opened up.

December 28, 2025 at 12:19 AM

Dan Sylveste

@constantinejn.bsky.social

“Quantum Advantage“

This is a graph plotting logical qbit lifetime in hours vs the year. The graph has multilines, all zero except for the near future which they go exponential. At the to is some text from the post:

“Because error corrected qubits recently begun to surpass physical qubits, I think the FOOM for logical qubits is near. I realize how ridiculous the hypotheticals look in this plot, but I legitimately think this is what will ‘

December 26, 2025 at 10:12 PM

Dan Sylveste

@constantinejn.bsky.social

The LOTR "aye, I could do that" meme with the following meta dialogue:

NEVER THOUGHT I'D BE -

HOW ABOUT I HELP YOU FRAME THAT THOUGHT CONTEXTUALLY

AYE, I COULD DO THAT

December 21, 2025 at 5:10 PM

Dan Sylveste

@constantinejn.bsky.social

Pretty good paper, frankly.

"This entire debate lacked basic mathematical scaffolding," Wolpert says.
"Once you build that scaffolding, the problem becomes clearer - and far more interesting."
At the core of his approach is a shift in perspective: instead of treating universes as physical systems with unknowable inner workings, treat them as kinds of computers. This lets Wolpert ground his model in the physical Church-Turing thesis, which holds that any physical process we can observe could, in principle, be reproduced by a standard computer program. Seen through this lens, the simulation question becomes a computational one - and mathematics, rather than speculation, sets the boundaries of what is possible.
That computational framing allows Wolpert to draw on a classical result from computer science known as Kleene's second recursion theorem, which explains how a program can generate and run an exact copy of itself. When Wolpert extends this theorem to entire universes, a surprising implication follows: if some universe can simulate ours accurately, nothing prevents our
universe from simulating that universe in return. Under certain assumptions, the two become mathematically indistinguishable, erasing the familiar hierarchy of "higher" and "lower" realities.
The framework also challenges a popular belief that deeper levels of simulation must be computationally weaker than the levels above them — an argument often used to claim that such chains must eventually terminate.
Wolpert shows that this isn't required by the mathematics: simulations do not have to degrade, and infinite chains of simulated universes remain fully consistent within the theory.

December 21, 2025 at 2:23 PM

Dan Sylveste

@constantinejn.bsky.social

What computer science has to say about the simulation hypothesis

iopscience.iop.org/article/10.1...

keywords:%20Church-Turing%20thesis,%20simulation%20hypothesis,%20computer%20science%20theory,%20the%20recursion%20theorem,%20Rice's%20theorem%0AAbstract%0AThe%20simulation%20hypothesis%20has%20recently%20excited%20renewed%20interest%20in%20the%20physics%20and%20philosophy%20communities.%20However,%20the%20hypothesis%20specifically%20concerns%20computers%20that%20simulate%20physical%20uni-verses.%20So%20to%20formally%20investigate%20the%20hypothesis,%20we%20need%20to%20understand%20it%20in%20terms%20of%20computer%20science%20(CS)%20theory.%20In%20addition%20we%20need%20a%20formal%20way%20to%20couple%20CS%20theory%20with%20physics.%20Here%20I%20couple%20those%20fields%20by%20using%20the%20physical%20Church-Turing%20thesis.%20This%20allow%20me%20to%20exploit%20Kleene's%20second%20recursion,%20to%20prove%20that%20not%20only%20is%20it%20possible%20for%20us%20to%20be%20a%20simulation%20being%20run%20on%20a%20com%20puter,%20but%20that%20we%20might%20be%20in%20a%20simulation%20that%20is%20being%20run%20on%20a%20computer%20-%20by%20us.%20In%20such%20a%20'self-simulation',%20there%20would%20be%20two%20identical%20instances%20of%20us,%20both%20equally%20real.%20I%20then%20use%20Rice's%20theorem%20to%20derive%20impossibility%20results%20concerning%20simulation%20and%20self-simulation;%20derive%20implication%20for%20(self-)simulation%20if%20we%20are%20being%20simulated%20in%20a%20program%20using%20fully%20homomorphic%20encryption;%20and%20briefly%20investigate%20the%20graphical%20structure%20of%20universes%20simulating%20other%20universes%20which%20contain%20computers%20running%20their%20own%20simulations.%20I%20end%20by%20describing%20some%20of%20the%20possible%20avenues%20for%20future%20research.%20While%20motivated%20in%20terms%20of%20the%20simulation%20hypothesis,%20the%20results%20in%20this%20pape%20are%20direct%20consequences%20of%20the%20Church-Turing%20thesis.%20So%20they%20apply%20far%20more%20broadly%20than%20the%20simulation%20hypothesis.%0AHe%20did%20not%20know%20if%…

December 21, 2025 at 2:23 PM

Dan Sylveste

@constantinejn.bsky.social

Don’t worry, I just be microdosing causality over here

arxiv.org/abs/2512.16898

Abstract In gauge theories, globally charged observables necessarily depend non-locally on the kinematical fields, with this dependence extending to the asymptotic boundary of spacetime. Despite this, we show that a subset of such observables can be consistently regarded as local to the bulk, in a manner that respects microcausality and leaves locality properties of uncharged observables untouched. A sufficient condition for this is to impose kinematically non-local boundary conditions on the large gauge sector of the theory, and to invoke a relational notion of localisation for observables. This reveals a relatively underappreciated link between boundary conditions, and different notions of microcausality and locality. We develop this point through a detailed case study in scalar QED, describing non-local boundary conditions that allow a large family of observables on a codimension-1 bulk surface to be viewed as local to that surface, despite being dressed by asymptotic Wilson lines. We show that this property continues to hold within a perturbative quantisation of the theory, and we argue that this leads to a consistent local net of algebras that includes these charged observables in bulk algebras. We explain how this setup may be understood in terms of a preterred
dynamical reference frame for small gauge transformations appearing in the boundary conditions. Many features of the theory (such as microcausality, the vacuum state, and the net of algebras of observables) depend on the choice of this frame, and we briefly discuss some repercussions of this for algebraic formulations of QFT. While our analysis is performed in QED, we expect our results
to carry over qualitatively to more complicated theories including gravity.

December 21, 2025 at 1:21 PM

Dan Sylveste

@constantinejn.bsky.social

Really feel like you never, ever hear about the potential downsides of targetable gigawatt masers directed from space, controlled by private interests accountable to no one.

I mean, we’ve all seen the James Bond movies, sure. But are these films a substitute for the discourse we should have?