lingo.lol is one of the many independent Mastodon servers you can use to participate in the fediverse.
A place for linguists, philologists, and other lovers of languages.

Server stats:

70
active users

#trust

5 posts5 participants0 posts today

AIs as Trusted Third Parties

This is a truly fascinating paper: “Trusted Machine Learning Models Unlock Private Inference for Problems Currently Infeasible with Cryptography.” The basic idea is that AIs can act as tr... schneier.com/blog/archives/202

Schneier on Security · AIs as Trusted Third Parties - Schneier on SecurityThis is a truly fascinating paper: “Trusted Machine Learning Models Unlock Private Inference for Problems Currently Infeasible with Cryptography.” The basic idea is that AIs can act as trusted third parties: Abstract: We often interact with untrusted parties. Prioritization of privacy can limit the effectiveness of these interactions, as achieving certain goals necessitates sharing private data. Traditionally, addressing this challenge has involved either seeking trusted intermediaries or constructing cryptographic protocols that restrict how much data is revealed, such as multi-party computations or zero-knowledge proofs. While significant advances have been made in scaling cryptographic approaches, they remain limited in terms of the size and complexity of applications they can be used for. In this paper, we argue that capable machine learning models can fulfill the role of a trusted third party, thus enabling secure computations for applications that were previously infeasible. In particular, we describe Trusted Capable Model Environments (TCMEs) as an alternative approach for scaling secure computation, where capable machine learning model(s) interact under input/output constraints, with explicit information flow control and explicit statelessness. This approach aims to achieve a balance between privacy and computational efficiency, enabling private inference where classical cryptographic solutions are currently infeasible. We describe a number of use cases that are enabled by TCME, and show that even some simple classic cryptographic problems can already be solved with TCME. Finally, we outline current limitations and discuss the path forward in implementing them...

Old friends coming over. We know each other since school, decades ago. We last met 7 years ago.

At my stage in life, this means wrinkles and all sorts of rather visible changes to our bodies. Not incapacity, necessarily, but an annoying dominance of wear and tear, all over.

Old friends meet, each absorbs the shock of seeing what is happening to the other, each recognises that their friend has the same experience.

Also: the joy of complete trust when meeting those old friends. Laughing together about subtle references to events 50 years back. The affirmation of having done a few things right in our lives. The easy recognition of past mistakes. All is good now. We have left those grudges and embarrassments behind us.

My braised ox cheek, slow-cooked in the chicken stock that I had made especially for the occasion in the week before: a success. Enjoying simple things together, based on a half century of shared experiences, good and bad.

Continued thread

Here's Chris Knight's key article on #ritual/speech coevolution as matrix and scaffold of #trust across #speech #communites. In 1998 he posited 130,000 years, and today it needs pushing back, coinciding with our process of speciation. At least the Chomsky crew here are conceding earlier dates for the critical transition to #language (they used to argue it was in the region 50-70,000 years wgo, which is now looking dumb).

#ritualspeechcoevolution #cooperation #communication

researchgate.net/publication/2

From my reading about social media platforms growing exponentially, there comes a point where growing a #moderation team is essential. This problem is less common with the #Fediverse with self-hosted small instances. But what about the larger instances? Have any #devs thought about implementing features such as is available in #Discourse. A #trust based moderation system as a community grows? Is that even a feasible thing to do?

🎙️✨Attention, Attention!
There’s a brand-new On Location podcast episode on ITSPmagazine

Interoperability, #AI, and the Future of #Trust in #Healthcare – A HIMSS 2025 Conversation

#Technology is transforming healthcare at an incredible pace—but as we digitize patient data, automate decision-making, and integrate AI into medical systems, we have to ask: Are we improving care, or just creating new risks?

In this virtual coverage of #HIMSS 2025, Sean Martin and I sat down with Robert Havasy to discuss interoperability, AI, and the evolving landscape of trust in healthcare. The healthcare industry is facing a critical moment: data needs to move seamlessly between systems, but #security, #privacy, and ethical concerns can’t be afterthoughts.

We dive into:
Interoperability – The promise and challenges of frictionless data exchange in healthcare
AI in healthcare – Can we trust machine-driven diagnostics and automated decision-making?
#Cybersecurity risks – As medical data moves faster, are we creating more vulnerabilities?
The societal impact – How does technology reshape our fundamental trust in healthcare?

We’re living in a time where convenience often wins over caution, but when it comes to our health, #privacy, and security, we can’t afford to take shortcuts. How do we ensure technology enhances human care rather than replacing it?

📺 Watch on YouTube: youtu.be/vufUNhxf9_M

🎧 Listen now: eventcoveragepodcast.com/episo

▶️ More On Location episodes: itspmagazine.com/technology-cy

🎥 ITSPmagazine YouTube Channel: youtube.com/@itspmagazine📺 Watch on YouTube: youtu.be/vufUNhxf9_M

🎧 Listen now: eventcoveragepodcast.com/episo

▶️ More On Location episodes: itspmagazine.com/technology-cy

🎥 ITSPmagazine YouTube Channel: youtube.com/@itspmagazine