Sal

Abstract

Although artificial intelligence enables productivity gains from delegating tasks to machines, it may facilitate the delegation of unethical behaviour. This risk is highly relevant amid the rapid rise of ‘agentic’ artificial intelligence systems. Here we demonstrate this risk by having human principals instruct machine agents to perform tasks with incentives to cheat. Requests for cheating increased when principals could induce machine dishonesty without telling the machine precisely what to do, through supervised learning or high-level goal setting. These effects held whether delegation was voluntary or mandatory. We also examined delegation via natural language to large language models. Although the cheating requests by principals were not always higher for machine agents than for human agents, compliance diverged sharply: machines were far more likely than human agents to carry out fully unethical instructions. This compliance could be curbed, but usually not eliminated, with the injection of prohibitive, task-specific guardrails. Our results highlight ethical risks in the context of increasingly accessible and powerful machine delegation, and suggest design and policy strategies to mitigate them.

This brainstorm is motivated by two recent discussions:

When is it time to switch away from youtube?

Lemmy community feedback: Should we fetch and combine comments from cross-posts on the post screen? · Issue #3415 · LemmyNet/lemmy-ui

The conversations about YouTube alternatives and PeerTube’s limitations made me test the current state of Lemmy <--> PeerTube federation.

Federation already works quite well: you can pull in PeerTube channels, subscribe, upvote, and comment from Lemmy.

But I did notice one thing. Take this Techlore video on PeerTube:

https://techlore.tv/w/9a7d20d2-4cc1-4911-8fd7-0794d8fe3bd9

-> On PeerTube: 6 upvotes, 4 comments

I pulled the video into my instance and was able to upvote it from there, so federated voting is working well: https://mander.xyz/post/37959974

I then noticed it had also been cross-posted to lemmy.world: https://lemmy.world/post/35205985

-> On lemmy.world: 37 upvotes, 9 comments

So the cross-post has significantly more engagement than the original video. This highlights a visibility issue: PeerTube creators may be getting more reach than they realize, but that engagement is fragmented across federated posts.

Brainstorm

Would it make sense to allow cross-posting of already-federated PeerTube content without breaking the link to the original object?

Upvotes/likes: These could be funneled back into the original PeerTube video, so creators get a true sense of their reach.

Comments: Less clear. Should all comments be merged into the video thread, or should they remain scoped to the Lemmy community they were posted in?

I don’t think this is “GitHub issue” ready, but it seems worth discussing.

The instance was updated today to v0.19.13

The official release notes can be found here: https://mander.xyz/post/37685278

The more noticeable frontend changes are:

Frontend

Don’t show edit mark if comment was edited in less than 5 minutes by @jfaustino #3197

Increase bio max length to 1000 chars by @nutomic #3249

Change link from element.io to matrix.org by @nutomic #3250

Remove all caches (fixes #3195) by @Nutomic in #3248

Fixed ordering for search results by @Nutomic in #3219

Add search field to community sidebar by @Nutomic in #3217

Add checkbox for title only search by @Nutomic in #3220

I decided to purchase a one-way pager, a programmer, and a paging subscription to satisfy my curiosity about pagers.

In this post, I am explaining my thought process and describing some of what I have learned about how pagers work. This is especially relevant to the national paging network in the Netherlands, but hopefully others also find it interesting.

The cellular network

Cellphones give us the ability to reach others and to remain reachable regardless of our location if within a network's coverage. The network infrastructure is continuously evolving in ways that make it more efficient, secure, and reliable.

One way that the network becomes more efficient is by improving its device tracking abilities to reduce the amount of radio broadcasting resources needed to deliver data to the recipient. Security and reliability are improved by having two-way communication between the network and devices such that devices can be authenticated, data correctly encrypted, and message delivery confirmed.

A participant within this network must accept one or more of their device's unique identifiers (at the very least the IMSI, often also the IMEI) is associated with an approximate location.

Since I do not want to accept these terms, I do not carry a phone with a SIM card on me.

A burner phone and an emergency pre-paid SIM card gives me the opportunity to connect to the network in the case that I need to contact someone immediately.

However, this does not give the opportunity to others to reach me in the case that they need me or worry about me. This is not common, but there have been cases in which being reachable would have been good.

LoRa / Meshtastic

Last year I learned about LoRa radios and the Meshtastic network implementation. These devices allow one to send encrypted messages directly between devices. The range is decent, especially if there is a line-of-sight between devices. With Meshtastic it is possible to create a network of nodes that route messages, and to make use of tunnels over the internet to connect nodes that are very far apart.

So far, my favorite use-cases for Meshtastic are communicating with my partner as I approach an area to meet them, communication during festivals/events, and when travelling in a small town or camping.

It is a great tool in some contexts, but I cannot be reliably reached with it.

The Pager

I am currently living in the Netherlands and so what I say is most relevant to the Dutch paging network 'KPN Nationaal 3'. Messages are broadcast using POCSAG 1200 at 172.450 MHz. I know that the situation with paging networks vary across the world, with paging networks being no longer available in many countries, but I don't know the details. It may be that the system here is rather special and unique.

The paging network is considered a legacy broadcasting system. Messages to the network are broadcast by transmitters distributed across the full coverage range. The message that is broadcast contains the RIC (Receiver Identify Code) and the message in plain text.

Anyone with an SDR (Software Defined Radio) device can decode and log all of the unencrypted messages. Here is an example using SDRConnect + multimon-ng:

Using a programming interface, a user can select the RIC codes that they want their network-tuned pager to be responsive to. The pager will beep and display on the screen messages sent to that RIC. In my case, the seller of the pager assigned a new RIC from their pool to me and programmed the pager to listen to it.

A pager does not have a built-in transmitter, and so it does not reveal any information to the network.

A subscription to the paging network works the following way:

• You get assigned your own 'RIC', which is publicly broadcast with every message
• You get assigned a private number (0665xxxxxx)
• While your subscription is active, you send an SMS or an e-mail to a specific address with your private number + message, and the network provider will broadcast it with the RIC as the recipient.

Then, anyone who knows your private number is able to reach a pager listening to your RIC. The public RIC is not enough information to request a message to be sent to you.

Registering to the network has a monthly cost (typical current pricing of 8 € - 20 €) depending on whether you want to be able to recieve text messages, numeric messages, or only make the pager beep. Your identity and banking information are known to the network provider. I was able to register as an individual without needing to provide any company information. I had to fill-in a short form and send it over e-mail with a photo of an ID to register.

So:

• The network provider knows your identity
• The service has a monthly cost
• The unencrypted message content, when they are sent, and the recipient's RIC are public information
• The network does not confirm delivery
• Inefficient for the network (all transmitters broadcast every message)
• Being a legacy system, the network may not remain alive for too long

But:

• It is possible to reach you at all times without needing to broadcast your location to the network

The pager is a technology that I looked at early on when I started thinking about privacy and I quickly discarded the idea. Giving my identity to a network provider and broadcasting unencrypted messages publicly did not seem logical to me.

Today, I see the value of having a receive-only device that is supported by a network with national coverage. A paging message would contain only enough information for me to know how urgently I need to find a way to communicate - whether I need to activate the burner phone immediately, or whether I can spend some time to go find another way to communicate.

For me, it was a pleasant surprise to discover that this legacy system fills the specific gap of reachability without tracking.

I also recently became aware of the existence of paging networks that rely on volunteer HAM radio operators (like DAPNET), and would like to explore these systems in the future.

Abstract

The mid-infrared (MIR) photonics market is rapidly expanding, driven by advancements in fiber-based MIR devices, particularly fiber lasers. However, the lack of robust MIR optical fibers remains a critical barrier to further technological progress. In this work, we present the fabrication of gallate glasses containing tantalum oxide, as it stands, the most robust mid-infrared glasses capable of being easily shaped into large bulk components, fibers or tapers. By introducing up to 20 mol% of tantalum oxide in a gallate glass, we achieve a two-order-of-magnitude improvement in water corrosion resistance, while the nonlinear refractive index increases tenfold compared to silica. Optimal thermal stability is attained at 10 mol% of tantalum oxide, enabling the fabrication of tens-of-meter-long optical fibers. Crucially, the addition of tantalum oxide enhances the gallate glass properties without compromising thermomechanical performance. The potential of these tantalo-gallate glasses is further demonstrated through supercontinuum generation in a laser-inscribed waveguide and a tapered fiber, spanning from the visible to 4.5 μm. This work establishes our developed tantalo-gallate glasses as a compelling alternative for photonic applications seeking robust mid-infrared materials, with the potential to overcome the critical barriers currently limiting the advancement of fiber-based MIR technologies.

Abstract

Understanding the radiative decay of exciton-polaritons is essential for achieving long-lived polaritons - a key prerequisite for enhancing nonlinear and quantum polaritonic effects. However, conventional wisdom - the coupled oscillator model - often oversimplifies polariton radiation as independent emissions from uncoupled excitonic and photonic resonances, overlooking the role of strong exciton-photon coupling in reshaping their radiative behavior. In this work, we present a theoretical framework that goes beyond the conventional coupled oscillator model by fully accounting for the collective and coherent nature of exciton-photon interactions. We demonstrate that these interactions can strongly suppress polariton radiation via destructive interference - both within the excitonic ensemble and between excitonic and photonic radiation channels - giving rise to polaritonic bound states in the continuum with infinitely long radiative lifetimes. Our approach offers a unified description of polariton radiative decay and establishes new design principles for engineering long-lived exciton-polaritons with tailored radiation properties, opening new avenues for nonlinear, topological, and quantum polaritonic applications. 

Abstract

Optical phase modulation is essential for a wide range of silicon photonic integrated circuits used in communication applications. In this study, an optical phase shifter utilizing photo-elastic effects is proposed, where mechanical stress is induced by electrostatic micro-electro-mechanical systems (MEMS) with actuators arranged in a comb drive configuration. The design incorporates suspended serpentine silicon nitride (SiN) optical waveguides. Through extensive numerical simulations, it is shown that the change in the effective refractive index (neff) of the optical waveguide is a function of the voltage applied to the electrostatic actuators and that such neff tuning can be achieved for a broad range of wavelengths. Implemented within one arm of an unbalanced Mach–Zehnder interferometer (MZI), the phase shifter achieves a phase change of π when the stressed optical path measures 4.7 mm, and the actuators are supplied with 80 V DC and consume almost no power. This results in a half-wave voltage-length product (VπL) of 37.6 V·cm. Comparative analysis with contemporary optical phase shifters highlights the proposed design’s superior power efficiency, compact footprint, and simplified fabrication process, making it a highly efficient component for reconfigurable MEMS-based silicon nitride photonic integrated circuits.

Abstract

Animals often sustain injuries, which are susceptible to lethal infections. In social insects, wound care behaviours have evolved to reduce these risks. But the limits of wound care behaviours remain unclear. Here we investigated the wound care behaviours of the ant Camponotus maculatus. Our findings show that amputation of legs infected with Pseudomonas aeruginosa significantly reduced mortality. However, nestmates do not differentiate between infected and sterile injuries, providing similar treatments regardless of infection. Even though we show that early amputation correlates with higher survival rates, nestmates amputate indiscriminately on legs with fresh or old wounds. Additionally, cuticular hydrocarbon profiles differed between ants with infected or sterile wounds only 24 hours post-injury, a timepoint when amputations are no longer effective. We propose that C. maculatus workers perform prophylactic amputations regardless of injury state or age. This is in sharp contrast to previous studies which showed clear capabilities to treat infected wounds differently in ants using antimicrobial compounds. This work therefore shows the limits of wound care behaviours in social insects, allowing us to better understand the evolutionary drivers of this unique behaviour.

Abstract

Established by the 1971 United Nations (UN) Convention on Psychotropic Substances, the prohibition of the recreational use of psychedelics (lysergic acid diethylamide [LSD], psilocybin, N,N-dimethyltryptamine [N,N-DMT], and mescaline) has two premises. First, recreational use poses a serious threat to public health because psychedelics are highly liable to addiction and abuse. Second, psychedelics have only limited scientific and medical uses. In this article, we raise the following questions: are these premises true such that prohibition is justified? If not, are decriminalization and legal regulation justified alternatives? Drawing on interdisciplinary research, we show that the premises of prohibition are false. Psychedelics are not highly liable to addiction or abuse, and so recreational use is not a serious threat to public health. Moreover, the uses of psychedelics exceed medical and scientific uses. Prohibition, we conclude, is therefore unjustified. We then show that decriminalization is based on the same false premises as prohibition, that legal regulation is based on weaker versions of these premises, and thus that both alternatives entail unjustified restrictions on recreational users. Finally, we present a fourth approach: communalization. This entails that all adults have the freedom to recreationally use psychedelics without restrictions and that communities provide harm reduction and benefit enhancement services to support this freedom.

For reference, you can find the 1971 UN document here: https://www.unodc.org/pdf/convention_1971_en.pdf

Abstract

The evolution of adaptive innovations carries strong eco-evolutionary implications, allowing organisms to explore novel ecological opportunities, which facilitates lineage diversification. The remarkable diversity of reproductive strategies in amphibians provides a natural laboratory for identifying ecological mechanisms driving evolutionary novelties. In viviparous salamanders, the transition from larviparous (i.e., live bearing of aquatic larvae) to pueriparous (i.e., live bearing of fully terrestrial juveniles) reproduction is hypothesized to represent an adaptation to the absence of water for larval deposition, in what is known as the dry-climate hypothesis. This work aimed to identify the ecological drivers of independent evolutionary transitions to pueriparity and test the dry-climate hypothesis.

Main Conclusions

Reproductive transitions to pueriparity in salamanders were consistently associated with scarcity of surface water driven by steep topography and karstic geology, providing the most unambiguous support for the dry-climate hypothesis to date and supporting convergent evolution of terrestrialisation in salamanders for most pueriparous lineages. Climatic, hydrological (i.e., soil moisture and vapour pressure deficit) and habitat factors appear to be comparatively less relevant drivers of reproductive shifts, yet may have played a role in specific transitions to pueriparity, especially at the intraspecific level. We also found key differences in the use of available habitat between reproductive modes, with pueriparity representing a more specialised strategy likely restricted to areas with strong selective pressure against aquatic breeding.

Abstract

Nonenzymatic assembly of RNA from chemically activated building blocks, such as phosphorimidazolides, would have been essential for the emergence of ribozymes on the early Earth. We previously showed that ribonucleoside monophosphates can be activated to phosphorimidazolides via a potentially prebiotic phospho-Passerini reaction involving 2-aminoimidazole, 2-methylbutyraldehyde, and methyl isocyanide, and that these activated nucleotides enable template-directed nonenzymatic RNA polymerization in the same reaction mixture. Here, we demonstrate that the same chemistry activates oligoribonucleotides and drives both nonenzymatic and ribozyme-catalyzed RNA ligation within the same reaction environment. By demonstrating a continuous path from prebiotic activation chemistry to RNA template copying by both nonenzymatic and ribozyme-catalyzed ligation, our results provide a more integrated and realistic model for primordial RNA assembly.

Abstract

 Spin-lattice coupling is crucial for understanding the spin transport and dynamics for spintronics and magnonics applications. Recently, cobalt titanate (CoTiO3), an easy-plane antiferromagnet, has been found to host axial phonons with a large magnetic moment, which may originate from spin-lattice coupling. Here, we investigate the effect of light-driven lattice dynamics on the magnetic properties of CoTiO3 using time-resolved spectroscopy with a THz pump and a magneto-optic probe. We found resonantly driven Raman active phonons, phonon-polariton-induced excitation of the antiferromagnetic magnons, and a slow increase in the polarization rotation of the probe, all indicating symmetry breaking that is not intrinsic to the magnetic space group. The temperature dependence confirmed that the observed spin dynamics is related to the magnetic order, and we suggest surface effects as a possible mechanism. Our results of THz-induced spin-lattice dynamics signify that extrinsic symmetry breaking may contribute strongly and unexpectedly to light-driven phenomena in bulk complex oxides.