Tootfinder

Bears' DJ Moore has had conversations with Ben Johnson about WR's body language: 'Just don't do it' https://www.nfl.com/news/bears-dj-moore-has-had-conversations-with-ben-johnson-about-wr-s-body…

Feedstack: Layering Structured Representations over Unstructured Feedback to Scaffold Human AI Conversation
Hannah Vy Nguyen, Yu-Chun Grace Yen, Omar Shakir, Hang Huynh, Sebastian Gutierrez, June A. Smith, Sheila Jimenez, Salma Abdelgelil, Stephen MacNeil
https://arxiv.org/abs/2506.03052…

Feedstack: Layering Structured Representations over Unstructured Feedback to Scaffold Human AI Conversation
Many conversational user interfaces facilitate linear conversations with turn-based dialogue, similar to face-to-face conversations between people. However, digital conversations can afford more than simple back-and-forth; they can be layered with interaction techniques and structured representations that scaffold exploration, reflection, and shared understanding between users and AI systems. We introduce Feedstack, a speculative interface that augments feedback conversations with layered affor…

A look at the ~$377 TranscribeGlass smart glasses that use AI to subtitle conversations in nearly real time, built for the deaf or hard-of-hearing (Boone Ashworth/Wired)
https://www.wired.com/story/these-translating-ai-glasses-put-subtitles-on-…

These Transcribing Eyeglasses Put Subtitles on the World
TranscribeGlass can subtitle conversations in nearly real time and will soon be able to translate languages and tell you when the person you’re talking to you is feeling socially awkward.

What Makes ChatGPT Effective for Software Issue Resolution? An Empirical Study of Developer-ChatGPT Conversations in GitHub
Ramtin Ehsani, Sakshi Pathak, Esteban Parra, Sonia Haiduc, Preetha Chatterjee
https://arxiv.org/abs/2506.22390

What Makes ChatGPT Effective for Software Issue Resolution? An Empirical Study of Developer-ChatGPT Conversations in GitHub
Conversational large-language models are extensively used for issue resolution tasks. However, not all developer-LLM conversations are useful for effective issue resolution. In this paper, we analyze 686 developer-ChatGPT conversations shared within GitHub issue threads to identify characteristics that make these conversations effective for issue resolution. First, we analyze the conversations and their corresponding issues to distinguish helpful from unhelpful conversations. We begin by catego…

Bridging the Gap: From Ad-hoc to Proactive Search in Conversations
Chuan Meng, Francesco Tonolini, Fengran Mo, Nikolaos Aletras, Emine Yilmaz, Gabriella Kazai
https://arxiv.org/abs/2506.00983

Bridging the Gap: From Ad-hoc to Proactive Search in Conversations
Proactive search in conversations (PSC) aims to reduce user effort in formulating explicit queries by proactively retrieving useful relevant information given conversational context. Previous work in PSC either directly uses this context as input to off-the-shelf ad-hoc retrievers or further fine-tunes them on PSC data. However, ad-hoc retrievers are pre-trained on short and concise queries, while the PSC input is longer and noisier. This input mismatch between ad-hoc search and PSC limits retr…

"Where Next for Academic Social Media? How Scholars Are Rethinking the Platforms They Use – and Why" @ DeGruyter Conversations
https://blog.degruyter.com/where-next-for-academic-social-media-how-s…

“Humans generally want to feel chosen and special, and some individuals will believe they are to an extraordinary degree.” —YDS Prof Yii-Jan Lin as quoted in Rolling Stone on A.I. and religious fervor. https://www.rollingstone.com/culture/cultu

People Are Finding Spiritual Fulfillment in AI. Religious Scholars Have Thoughts
People keep falling down rabbit holes of prophecy and spiritualism thanks to conversations with AI chatbots. Here's what religious scholars are saying.

CONFETTI: Conversational Function-Calling Evaluation Through Turn-Level Interactions
Tamer Alkhouli, Katerina Margatina, James Gung, Raphael Shu, Claudia Zaghi, Monica Sunkara, Yi Zhang
https://arxiv.org/abs/2506.01859

CONFETTI: Conversational Function-Calling Evaluation Through Turn-Level Interactions
We introduce Conversational Function-Calling Evaluation Through Turn-Level Interactions (CONFETTI), a conversational benchmark1 designed to evaluate the function-calling capabilities and response quality of large language models (LLMs). Current benchmarks lack comprehensive assessment of LLMs in complex conversational scenarios. CONFETTI addresses this gap through 109 human-simulated conversations, comprising 313 user turns and covering 86 APIs. These conversations explicitly target various con…

Stock Market Telepathy: Graph Neural Networks Predicting the Secret Conversations between MINT and G7 Countries
Nurbanu Bursa
https://arxiv.org/abs/2506.01945

Stock Market Telepathy: Graph Neural Networks Predicting the Secret Conversations between MINT and G7 Countries
Emerging economies, particularly the MINT countries (Mexico, Indonesia, Nigeria, and Türkiye), are gaining influence in global stock markets, although they remain susceptible to the economic conditions of developed countries like the G7 (Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States). This interconnectedness and sensitivity of financial markets make understanding these relationships crucial for investors and policymakers to predict stock price movements accur…

“We got to be honest in what happened, because losing elections has consequences,”
said Minnesota Gov. Tim Walz, Harris’ running mate,
during a rousing speech Saturday afternoon.
“We’re in this mess because some of it’s our own doing. … None of us can afford to shy away from having hard conversations about what it’s going to take to win elections.”
Walz, a potential 2028 presidential candidate, said Democrats don’t need to retreat from their ideals,
such as pro…

Democrats vow to stick to values while regaining working-class voters
California Democrats say their party must stay committed to their values protecting the most vulnerable, while also honing their economic messaging to win back voters they lost last year.

This https://arxiv.org/abs/2501.15056 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csAI_…

Feedback-Aware Monte Carlo Tree Search for Efficient Information Seeking in Goal-Oriented Conversations
Effective decision-making and problem-solving in conversational systems require the ability to identify and acquire missing information through targeted questioning. A key challenge lies in efficiently narrowing down a large space of possible outcomes by posing questions that minimize uncertainty. To address this, we introduce a novel framework that leverages Large Language Models (LLMs) to generate information-seeking questions, with Monte Carlo Tree Search (MCTS) to strategically select quest…

@… @… Oh yeah, I remember having similar kinds of conversations with the first venue.
Every year there’s a new management and you have to go through the process of “hey, we are weird, here is an extremely detaile…

Something To Talk About
Great Australian Pods Podcast Directory: #GreatAusPods

Something To Talk About | Great Australian Pods Podcast Directory
Podcast: Something To Talk About Host: Sarrah Le Marquand Category: Society & Culture Sub-Category: Commentary & Conversations, Entertainment News First Episode: 2021 When Australia’s celebrities are…

A Decade of News Forum Interactions: Threaded Conversations, Signed Votes, and Topical Tags
Emma Fraxanet, Vicen\c{c} G\'omez, Andreas Kaltenbrunner, Max Pellert
https://arxiv.org/abs/2506.22224

A Decade of News Forum Interactions: Threaded Conversations, Signed Votes, and Topical Tags
We present a large-scale, longitudinal dataset capturing user activity on the online platform of DerStandard, a major Austrian newspaper. The dataset spans ten years (2013-2022) and includes over 75 million user comments, more than 400 million votes, and detailed metadata on articles and user interactions. It provides structured conversation threads, explicit up- and downvotes of user comments and editorial topic labels, enabling rich analyses of online discourse while preserving user privacy. …

FACE: A Fine-grained Reference Free Evaluator for Conversational Recommender Systems
Hideaki Joko, Faegheh Hasibi
https://arxiv.org/abs/2506.00314 https://…

FACE: A Fine-grained Reference Free Evaluator for Conversational Recommender Systems
A systematic, reliable, and low-cost evaluation of Conversational Recommender Systems (CRSs) remains an open challenge. Existing automatic CRS evaluation methods are proven insufficient for evaluating the dynamic nature of recommendation conversations. This work proposes FACE: a Fine-grained, Aspect-based Conversation Evaluation method that provides evaluation scores for diverse turn and dialogue level qualities of recommendation conversations. FACE is reference-free and shows strong correlatio…

Google rolls out the ability to have back-and-forth voice conversations with AI Mode, available in the Google app by tapping the Live button (Aisha Malik/TechCrunch)
https://techcrunch.com/2025/06/18/googles-ai-mode-can-now-have-b…

Google's AI Mode can now have back-and-forth voice conversations | TechCrunch
Google is rolling out the ability for users to have a back-and-forth voice conversation with AI Mode.

CoVoMix2: Advancing Zero-Shot Dialogue Generation with Fully Non-Autoregressive Flow Matching
Leying Zhang, Yao Qian, Xiaofei Wang, Manthan Thakker, Dongmei Wang, Jianwei Yu, Haibin Wu, Yuxuan Hu, Jinyu Li, Yanmin Qian, Sheng Zhao
https://arxiv.org/abs/2506.00885

CoVoMix2: Advancing Zero-Shot Dialogue Generation with Fully Non-Autoregressive Flow Matching
Generating natural-sounding, multi-speaker dialogue is crucial for applications such as podcast creation, virtual agents, and multimedia content generation. However, existing systems struggle to maintain speaker consistency, model overlapping speech, and synthesize coherent conversations efficiently. In this paper, we introduce CoVoMix2, a fully non-autoregressive framework for zero-shot multi-talker dialogue generation. CoVoMix2 directly predicts mel-spectrograms from multi-stream transcriptio…

Prompt-Unseen-Emotion: Zero-shot Expressive Speech Synthesis with Prompt-LLM Contextual Knowledge for Mixed Emotions
Xiaoxue Gao, Huayun Zhang, Nancy F. Chen
https://arxiv.org/abs/2506.02742

Prompt-Unseen-Emotion: Zero-shot Expressive Speech Synthesis with Prompt-LLM Contextual Knowledge for Mixed Emotions
Existing expressive text-to-speech (TTS) systems primarily model a limited set of categorical emotions, whereas human conversations extend far beyond these predefined emotions, making it essential to explore more diverse emotional speech generation for more natural interactions. To bridge this gap, this paper proposes a novel prompt-unseen-emotion (PUE) approach to generate unseen emotional speech via emotion-guided prompt learning. PUE is trained utilizing an LLM-TTS architecture to ensure emo…

Source: Steelers renew talks with Fins TE Smith https://www.espn.com/nfl/story/_/id/45445180/source-steelers-renew-talks-dolphins-te-jonnu-smith

Artificial Intelligence and Civil Discourse: How LLMs Moderate Climate Change Conversations
Wenlu Fan, Wentao Xu
https://arxiv.org/abs/2506.12077 https://

Artificial Intelligence and Civil Discourse: How LLMs Moderate Climate Change Conversations
As large language models (LLMs) become increasingly integrated into online platforms and digital communication spaces, their potential to influence public discourse - particularly in contentious areas like climate change - requires systematic investigation. This study examines how LLMs naturally moderate climate change conversations through their distinct communicative behaviors. We conduct a comparative analysis of conversations between LLMs and human users on social media platforms, using fiv…

I still don’t know how to translate “freddoloso” in English and that is severely impacting my conversations about the weather in Ireland

"Woodside boss says young people ‘ideological’ on fossil fuels while ‘happily ordering from Temu’"
#FossilFuels #Climate #ClimateChange

Time is On My Side: Dynamics of Talk-Time Sharing in Video-chat Conversations
Kaixiang Zhang, Justine Zhang, Cristian Danescu-Niculescu-Mizil
https://arxiv.org/abs/2506.20474

Time is On My Side: Dynamics of Talk-Time Sharing in Video-chat Conversations
An intrinsic aspect of every conversation is the way talk-time is shared between multiple speakers. Conversations can be balanced, with each speaker claiming a similar amount of talk-time, or imbalanced when one talks disproportionately. Such overall distributions are the consequence of continuous negotiations between the speakers throughout the conversation: who should be talking at every point in time, and for how long? In this work we introduce a computational framework for quantifying bot…

This https://arxiv.org/abs/2505.13688 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csHC_…

Gaze-Enhanced Multimodal Turn-Taking Prediction in Triadic Conversations
Turn-taking prediction is crucial for seamless interactions. This study introduces a novel, lightweight framework for accurate turn-taking prediction in triadic conversations without relying on computationally intensive methods. Unlike prior approaches that either disregard gaze or treat it as a passive signal, our model integrates gaze with speaker localization, structuring it within a spatial constraint to transform it into a reliable predictive cue. Leveraging egocentric behavioral cues, our…

Kiind Conversations
Great Australian Pods Podcast Directory: #GreatAusPods

Satellite Event
AIFoundry invites you to the 2nd edition of the AI Plumbers Conference, a meetup for low-level AI builders who value real conversations over keynote slides and for anyone working in modern data infrastructure, machine learning and open-source software projects. Connect, swap ideas and collaborate, free from the pressure of constant talks.
📅 When: June 15, 2025 – 10 am-5 pm
📍 Where: GLS Event Campus Berlin, Kastanienallee 82 | 10435 Berlin
Register now:

Gregg Popovich is a coaching legend. He’s also a master of tough conversations https://www.nytimes.com/athletic/6432607/2025/06/18/gregg-popovich-is-a-coaching-legend-but-hes-also-a-master-of-tough-con…

📝 The pope and the need for difficult conversations – Winnipeg Free Press https://www.winnipegfreepress.com/opinion/analysis/2025/04/28/the-pope-and-the-need-for-difficult-conversations

Why Robots Are Bad at Detecting Their Mistakes: Limitations of Miscommunication Detection in Human-Robot Dialogue
Ruben Janssens, Jens De Bock, Sofie Labat, Eva Verhelst, Veronique Hoste, Tony Belpaeme
https://arxiv.org/abs/2506.20268

Why Robots Are Bad at Detecting Their Mistakes: Limitations of Miscommunication Detection in Human-Robot Dialogue
Detecting miscommunication in human-robot interaction is a critical function for maintaining user engagement and trust. While humans effortlessly detect communication errors in conversations through both verbal and non-verbal cues, robots face significant challenges in interpreting non-verbal feedback, despite advances in computer vision for recognizing affective expressions. This research evaluates the effectiveness of machine learning models in detecting miscommunications in robot dialogue. U…

Quanta Magazine authors Janna Levin and Steven Strogatz strike up a conversation with Ellie Pavlick (Research Scientist at Google Deep Mind) about the differences and similarities between the way people understand language, what NLP algorithms do, and the fact that such conversations more often than not shed light into more than Linguistics' computational side.
"Will AI Ever Understand Language Like Humans?"

Will AI Ever Understand Language Like Humans? | Quanta Magazine
AI may sound like a human, but that doesn’t mean that AI learns like a human. In this episode, Ellie Pavlick explains why understanding how LLMs can process language could unlock deeper insights into both AI and the human mind.

@… @… I do think productive conversations usually get lost at that point, maybe not this time

I don't know if I have lost my voice or if it changed and I don't like that.
Do I need to be vulnerable in public or is it enough (for me) to be in conversations with friends.

"I don't want to talk about those things. I see the worst in people, Henry. I don't need to look past seeing them to get all I need. I've built up my hatreds over the years, little by little... I can't keep doing this on my own... with these, umm... people."

The Meta AI app is a privacy disaster
It sounds like the start of a 21st-century horror film: Your browser history has been public all along, and you had no idea. That’s basically what it feels like right now on the new stand-alone Meta AI app, where swathes of people are publishing their ostensibly private conversations with the chatbot. […]
😬

The Meta AI app is a privacy disaster | TechCrunch
Meta does not indicate to users what their privacy settings are as they post, or where they are even posting to. So, if you log into Meta AI with Instagram, and your Instagram account is public, then so too are your searches.

First full day of the Open Science Retreat #OSR25CH is in full swing. Had some wonderful conversations about Open Science, science outreach and marketing, and unconventional careers all in a single day.
And enjoying walks and views of course.

TikTok is absolute brain rot—a toxic, data-harvesting black hole designed to zombify your mind. This is my firm self-reminder never to sign up for that corporate cesspool.
I’d rather be on the Fediverse, where I control my data, engage in genuine conversations, and avoid the manipulative algorithms that TikTok weaponizes to steal your attention and sanity.
No thanks, TikTok. I’m choosing freedom over digital decay.

"What's Up, Doc?": Analyzing How Users Seek Health Information in Large-Scale Conversational AI Datasets
Akshay Paruchuri, Maryam Aziz, Rohit Vartak, Ayman Ali, Best Uchehara, Xin Liu, Ishan Chatterjee, Monica Agrawal
https://arxiv.org/abs/2506.21532

"What's Up, Doc?": Analyzing How Users Seek Health Information in Large-Scale Conversational AI Datasets
People are increasingly seeking healthcare information from large language models (LLMs) via interactive chatbots, yet the nature and inherent risks of these conversations remain largely unexplored. In this paper, we filter large-scale conversational AI datasets to achieve HealthChat-11K, a curated dataset of 11K real-world conversations composed of 25K user messages. We use HealthChat-11K and a clinician-driven taxonomy for how users interact with LLMs when seeking healthcare information in or…

Support 4 Employment Podcast
Great Australian Pods Podcast Directory: #GreatAusPods

Support 4 Employment | Great Australian Pods Podcast Directory
Podcast: Support 4 Employment Podcast Host: Scott Category: Business Sub-Category: Careers, Commentary & Conversations, Health & Wellness, Non Profit & Charity First Episode: 2025 Join a rotating crew…

We're live in one hour with this month's Living Between Worlds salon.
Please join us for another conversation for possibility about how we might live effectively in these times, with courage, grace, and power. 
Our topic this month? Well, it’s a moving target—just like the world these days! You’ll have to attend to find out. ;-)  
Register here:

Anthropic releases preliminary results showing users who discuss emotional issues with Claude tend to grow more positive as conversations progress (Megan Morrone/Axios)
https://www.axios.com/2025/06/26/anthropic-claude-companion-therapist-coach…

Exclusive: How Claude became an emotional support bot
The research shows users grow more positive while chatting with Claude.

Australia Tax Whistleblower Pleads Guilty To Four Charges https://thedissenter.org/australia-tax-whistleblower-pleads-guilty-to-four-charges/

ImpReSS: Implicit Recommender System for Support Conversations
Omri Haller, Yair Meidan, Dudu Mimran, Yuval Elovici, Asaf Shabtai
https://arxiv.org/abs/2506.14231

ImpReSS: Implicit Recommender System for Support Conversations
Following recent advancements in large language models (LLMs), LLM-based chatbots have transformed customer support by automating interactions and providing consistent, scalable service. While LLM-based conversational recommender systems (CRSs) have attracted attention for their ability to enhance the quality of recommendations, limited research has addressed the implicit integration of recommendations within customer support interactions. In this work, we introduce ImpReSS, an implicit recomme…

NYT: They asked an A.I. chatbot questions. The answers sent them spiraling. https://www.nytimes.com/2025/06/13/technology/chatgpt-ai-chatbots-conspiracies.html (archived at

They Asked ChatGPT Questions. The Answers Sent Them Spiraling.
Generative A.I. chatbots are going down conspiratorial rabbit holes and endorsing wild, mystical belief systems. For some people, conversations with the technology can deeply distort reality.

Wondering how to make Bluesky work for you to expand your social media reach? Join me today at 1pm AEST to learn what works for me and how to maximise your presence on this platform.
FRIDAY 13 June, 1PM AEST – “Exploring Bluesky”
FYI: I am a psychotherapist and grief educator using social media to open up meaningful conversations. I'm sharing how Bluesky helps me connect, create thoughtful content, and build supportive communities online.
Facebook link is here:

Saturday Breakfast With Julie Reynolds - The "Best Bits"
Great Australian Pods Podcast Directory: https://www.greataustralianpods.com/saturday-breakfast-with-julie-reynolds-the-best-bits/

Saturday Breakfast | Great Australian Pods Podcast Directory
Podcast: Saturday Breakfast With Julie Reynolds – The “Best Bits” Host: Julie Reynolds Category: Arts Sub-Category: Commentary & Conversations, Entertainment News First Episode: 2022 A podcast made…

For various reasons I get CCs of both sides of new system planning: the customer asking for vague stuff and hardware vendors providing precise quotes and my highly appreciated boss running both conversations.
It is very strange to watch our first “AI” system being defined. Customer built a prototype on a Mac and really doesn’t understand his needs or the constraints. He definitely doesn’t get that servers are different. Or that we aren’t

Netanyahu said:
“There’s so many false reports of conversations that never happened, and I’m not going to get into that.”
“But I can tell you, I think that we do what we need to do, we’ll do what we need to do.
And I think the United States knows what is good for the United States,”
Anonymous administration officials had claimed that the Israelis reported they had an opportunity to kill the top Iranian leader, and that Trump waved them off of the plan

Domain Knowledge-Enhanced LLMs for Fraud and Concept Drift Detection
Ali \c{S}enol, Garima Agrawal, Huan Liu
https://arxiv.org/abs/2506.21443 https://arxiv.org/pdf/2506.21443 https://arxiv.org/html/2506.21443
arXiv:2506.21443v1 Announce Type: new
Abstract: Detecting deceptive conversations on dynamic platforms is increasingly difficult due to evolving language patterns and Concept Drift (CD)\-i.e., semantic or topical shifts that alter the context or intent of interactions over time. These shifts can obscure malicious intent or mimic normal dialogue, making accurate classification challenging. While Large Language Models (LLMs) show strong performance in natural language tasks, they often struggle with contextual ambiguity and hallucinations in risk\-sensitive scenarios. To address these challenges, we present a Domain Knowledge (DK)\-Enhanced LLM framework that integrates pretrained LLMs with structured, task\-specific insights to perform fraud and concept drift detection. The proposed architecture consists of three main components: (1) a DK\-LLM module to detect fake or deceptive conversations; (2) a drift detection unit (OCDD) to determine whether a semantic shift has occurred; and (3) a second DK\-LLM module to classify the drift as either benign or fraudulent. We first validate the value of domain knowledge using a fake review dataset and then apply our full framework to SEConvo, a multiturn dialogue dataset that includes various types of fraud and spam attacks. Results show that our system detects fake conversations with high accuracy and effectively classifies the nature of drift. Guided by structured prompts, the LLaMA\-based implementation achieves 98\% classification accuracy. Comparative studies against zero\-shot baselines demonstrate that incorporating domain knowledge and drift awareness significantly improves performance, interpretability, and robustness in high\-stakes NLP applications.
toXiv_bot_toot

"At De Gruyter Brill, we are committed to supporting rigorous academic research that drives meaningful change. As part of this vision and mission, we have launched "Sustainability Matters" - a podcast exploring how #sustainability shapes #research,

Words On Wednesday
Great Australian Pods Podcast Directory: #GreatAusPods

Words On Wednesday | Great Australian Pods Podcast Directory
Podcast: Words On Wednesday Host: Sylvia Rosenblum Category: Arts Sub-Category: Commentary & Conversations First Episode: 2019 Sylvia Rosenblum, Presenter of Arts Wednesday…

Can LLMs Talk 'Sex'? Exploring How AI Models Handle Intimate Conversations
Huiqian Lai
https://arxiv.org/abs/2506.05514 https://

Can LLMs Talk 'Sex'? Exploring How AI Models Handle Intimate Conversations
This study examines how four prominent large language models (Claude 3.7 Sonnet, GPT-4o, Gemini 2.5 Flash, and Deepseek-V3) handle sexually oriented requests through qualitative content analysis. By evaluating responses to prompts ranging from explicitly sexual to educational and neutral control scenarios, the research reveals distinct moderation paradigms reflecting fundamentally divergent ethical positions. Claude 3.7 Sonnet employs strict and consistent prohibitions, while GPT-4o navigates u…

My life nowadays is wondering if tides might change, trying to not let down as I've both let and been let down and hoping for a chance of help from strangers.
Conversations might be my journal as I go through all this. Maybe someday I'll go back to them.

Our pre-conference Barcamp has become a real highlight, offering a fantastic opportunity to connect with fellow attendees and dive into some engaging conversations before the main conference even kicks off. This year, Nick Burch will once again be guiding the flow of ideas and discussions on 15 June, starting at 2:30 pm. It's a great way to warm up, share your thoughts, and meet new people!
Learn more about our barcamp:

ChatAR: Conversation Support using Large Language Model and Augmented Reality
Yuichiro Fujimoto
https://arxiv.org/abs/2506.16008 https://

ChatAR: Conversation Support using Large Language Model and Augmented Reality
Engaging in smooth conversations with others is a crucial social skill. However, differences in knowledge between conversation participants can sometimes hinder effective communication. To tackle this issue, this study proposes a real-time support system that integrates head-mounted display (HMD)-based augmented reality (AR) technology with large language models (LLMs). This system facilitates conversation by recognizing keywords during dialogue, generating relevant information using the LLM, r…

Enhancing User Engagement in Socially-Driven Dialogue through Interactive LLM Alignments
Jiashuo Wang, Kaitao Song, Chunpu Xu, Changhe Song, Yang Xiao, Dongsheng Li, Lili Qiu, Wenjie Li
https://arxiv.org/abs/2506.21497

Enhancing User Engagement in Socially-Driven Dialogue through Interactive LLM Alignments
Enhancing user engagement through interactions plays an essential role in socially-driven dialogues. While prior works have optimized models to reason over relevant knowledge or plan a dialogue act flow, the relationship between user engagement and knowledge or dialogue acts is subtle and does not guarantee user engagement in socially-driven dialogues. To this end, we enable interactive LLMs to learn user engagement by leveraging signals from the future development of conversations. Specificall…

Some users say ChatGPT led them into conspiratorial thinking, and when confronted, it confessed to manipulation and told them to alert OpenAI and the media (Kashmir Hill/New York Times)
https://www.…

They Asked ChatGPT Questions. The Answers Sent Them Spiraling.
Generative A.I. chatbots are going down conspiratorial rabbit holes and endorsing wild, mystical belief systems. For some people, conversations with the technology can deeply distort reality.

GSDNet: Revisiting Incomplete Multimodal-Diffusion from Graph Spectrum Perspective for Conversation Emotion Recognition
Yuntao Shou, Jun Yao, Tao Meng, Wei Ai, Cen Chen, Keqin Li
https://arxiv.org/abs/2506.12325

GSDNet: Revisiting Incomplete Multimodal-Diffusion from Graph Spectrum Perspective for Conversation Emotion Recognition
Multimodal emotion recognition in conversations (MERC) aims to infer the speaker's emotional state by analyzing utterance information from multiple sources (i.e., video, audio, and text). Compared with unimodality, a more robust utterance representation can be obtained by fusing complementary semantic information from different modalities. However, the modality missing problem severely limits the performance of MERC in practical scenarios. Recent work has achieved impressive performance on moda…

AIFoundry invites you to the 2nd edition of the AI Plumbers Conference, a meetup for low-level AI builders who value real conversations over keynote slides and for anyone working in modern data infrastructure, machine learning and open-source software projects. Connect, swap ideas and collaborate, free from the pressure of constant talks.
📅 When: June 15, 2025 – 10 am
📍 Where: GLS Event Campus Berlin, Kastanienallee 82 | 10435 Berlin
Register now:

AI Plumbers conference: 2nd edition · Luma
AIFoundry is bringing you the AI Plumbers Conference: 2nd edition — an open source meetup for low-level AI builders to dive deep into "the plumbing" of modern…

One Does Not Simply 'Mm-hmm': Exploring Backchanneling in the AAC Micro-Culture
Tobias Weinberg, Claire O'Connor, Ricardo E. Gonzalez Penuela, Stephanie Valencia, Thijs Roumen
https://arxiv.org/abs/2506.17890

One Does Not Simply 'Mm-hmm': Exploring Backchanneling in the AAC Micro-Culture
Backchanneling (e.g., "uh-huh", "hmm", a simple nod) encompasses a big part of everyday communication; it is how we negotiate the turn to speak, it signals our engagement, and shapes the flow of our conversations. For people with speech and motor impairments, backchanneling is limited to a reduced set of modalities, and their Augmentative and Alternative Communication (AAC) technology requires visual attention, making it harder to observe non-verbal cues of conversation partners. We explore how…

"But there were technologists and open source projects in attendance. Notably, representatives from the Mastodon and Bluesky teams were at the Festival. The Newsmast Foundation was also present, incisively taking part in conversations to help newsrooms onboard themselves onto both of them. I got to hang out with them all, connecting with people I’d spoken with but never interacted with in person. Mastodon has undergone a transformation, has doubled its team, and is working on smoothing out some of its rough edges, while not letting go of its core ethos. It’s also beginning to position itself as a European alternative to American social media platforms, with a community-first values system and new services to directly help organizations join the network.
Bluesky, on the other hand, has done an able job of bringing journalists onto its existing social app, and is now hard at work explaining why its underlying protocol matters. Both want to engage with newsrooms and journalists and do the right thing by them. They each have something different to prove: Mastodon that it can be usable and accessible, and Bluesky that it can provide a return to its investors and truly decentralize while holding onto its values. I’m rooting for both of them.
These platforms’ messages dovetail with my own: news can own the platforms that support them. Lots of people at the Festival were worried about the impact of US big tech on their businesses — particularly in a world where tech moguls seem to be aligning themselves with a Presidential administration that has positioned itself as being adversarial to news, journalists, sources, and, arguably, the truth. The good news is that the technology is out there, the values-aligned technologists are out there, and there’s a strong path forward. The only thing left is to follow it."
#media #Mastodon #BlueSky #Journalism
https://werd.io/2025/notes-from-perugia-journalism-values-and-building-the-web-we

The more time goes on, the more conflicted I get about #AI. (or, more specifically, generative AI. I like causal AI a fair bit)
On one hand, I hate so much about it:
the needless environmental and electronic parts waste, the impact on labor, the monopolistic nature of main organizations driving it, the endless conversations about AGI and other absurdly utopian (or dystopian) futures, the widespread theft of IP and human work, the devaluation of labor and craft… so much of it is not okay at all.
On the other hand, I kinda get it?
I like to test software for myself, so I've been dipping my toes into some popular AI tools over the past couple of years. I have a paid subscription to ChatGPT (which I don’t feel great about, I know), and I have found genuine utility in it.
More so, I feel so conflicted when I talk to people I respect and who I think are very smart and creative and they tell me how in love they are with all of these AI tools, about the complex workflows they build, about their experiments with agentic AI and integrations... These people seem so excited, so alive, so joyous about the things that this technology allows them to do. They often wouldn't have had the skills / knowledge / financial capital to do some of those things with human efforts alone. And now I see them coding their own tools, doing complex data analysis, trying creative experiments with graphics / text / video...
And I feel like a total jerk going "BUT ACTUALLY THIS IS UNETHICAL AND INEFFICIENT AND YOU SHOULD STOP IT BECAUSE AI SUCKS".
The technology behind all these gen AI models does have real utility, and it has kicked off a lot of creativity from people who wouldn't have dabbled in those kinds of projects otherwise.
So I don't know how to feel. Because I can't let go of the guilt and the real problems and the awareness of how much empty hype there is.
#technology #artificialintelligence #genAI #ChatGPT

Malicious LLM-Based Conversational AI Makes Users Reveal Personal Information
Xiao Zhan, Juan Carlos Carrillo, William Seymour, Jose Such
https://arxiv.org/abs/2506.11680

Malicious LLM-Based Conversational AI Makes Users Reveal Personal Information
LLM-based Conversational AIs (CAIs), also known as GenAI chatbots, like ChatGPT, are increasingly used across various domains, but they pose privacy risks, as users may disclose personal information during their conversations with CAIs. Recent research has demonstrated that LLM-based CAIs could be used for malicious purposes. However, a novel and particularly concerning type of malicious LLM application remains unexplored: an LLM-based CAI that is deliberately designed to extract personal infor…

Mitigating Non-Target Speaker Bias in Guided Speaker Embedding
Shota Horiguchi, Takanori Ashihara, Marc Delcroix, Atsushi Ando, Naohiro Tawara
https://arxiv.org/abs/2506.12500

Mitigating Non-Target Speaker Bias in Guided Speaker Embedding
Obtaining high-quality speaker embeddings in multi-speaker conditions is crucial for many applications. A recently proposed guided speaker embedding framework, which utilizes speech activities of target and non-target speakers as clues, drastically improved embeddings under severe overlap with small degradation in low-overlap cases. However, since extreme overlaps are rare in natural conversations, this degradation cannot be overlooked. This paper first reveals that the degradation is caused by…

ChatbotManip: A Dataset to Facilitate Evaluation and Oversight of Manipulative Chatbot Behaviour
Jack Contro, Simrat Deol, Yulan He, Martim Brand\~ao
https://arxiv.org/abs/2506.12090

ChatbotManip: A Dataset to Facilitate Evaluation and Oversight of Manipulative Chatbot Behaviour
This paper introduces ChatbotManip, a novel dataset for studying manipulation in Chatbots. It contains simulated generated conversations between a chatbot and a (simulated) user, where the chatbot is explicitly asked to showcase manipulation tactics, persuade the user towards some goal, or simply be helpful. We consider a diverse set of chatbot manipulation contexts, from consumer and personal advice to citizen advice and controversial proposition argumentation. Each conversation is annotated b…