Tommaso Turchi

Chain-of-Thought Is Not Explainability — Paper Review Rethinking How We Understand AI Reasoning

Humans are terrible at explaining their own decision-making. We confabulate logical stories for intuitive choices. New research shows LLMs do the same - with serious implications for medicine, law, and other high-stakes applications.

Quality of scientific papers questioned as academics ‘overwhelmed’ by the millions published Widespread mockery of AI-generated rat with giant penis in one paper brings problem to public attention

It's not AI that's undermining science (or education, or so many other fields) - it's the relentless push for more output with little reward for quality. AI just amplifies a pre-existing problem by offering a quick, sloppy way out.

The Computer-Science Bubble Is Bursting Artificial intelligence is ideally suited to replacing the very type of person who built it.

This piece is a wake-up call: entry-level devs can't get the on-the-job training opportunities past generations had. AI eliminated the bottom rung of the ladder. Universities need to step up and provide the practical experience industry won't.

Can you spare 10–15 min for an #AI research project? My student is studying how explanations impact trust in AI decisions. No AI experience needed!

👉English: https://survey.trx.li/index.php/193548?lang=en
👉Italiano: https://survey.trx.li/index.php/193548?lang=it

RTs greatly appreciated! #Explainab

The illusion of conscious AI Neuroscientist Anil Seth explains why people often overestimate how likely it is that AI will become conscious.

“Nobody expects a computer simulation of a hurricane to generate real wind and real rain.” Neuroscientist Anil Seth argues we overestimate the odds of conscious AI—intelligence ≠ consciousness. We should be cautious about trying to create conscious machines. bigthink.com/neuropsych/t...

If you're in AI or education—or simply affected by them (so everyone at this point)—you should watch this Veritasium talk.

It connects Kahneman's System 1 & 2 to why "education revolutions" keep falling short, and what AI might change (for better or worse).

https://www.youtube.com/watch?v=0xS68sl2

Stop trying to make LLMs run your app logic. They're language models—use them to interpret what the user meant, then hand off to real code. Great piece: https://sgnt.ai/p/hell-out-of-llms/

TLDR: Get in, get meaning, get the hell out.

🎉 Wrapping up an incredible workshop on Adaptive eXplainable AI (#AXAI)! Huge thanks to everyone who joined, shared insights, and sparked exciting conversations. Already counting down to next year's edition—expect even more innovation and collaboration! Stay tuned! #XAI

Just landed in Cagliari! 📍 Tomorrow's the big day—AXAI Workshop is almost here. We've just uploaded the camera-ready papers of all accepted contributions. Check them out here: https://axai.trx.li/accepted-papers/ #AXAI #XAI #IUI2025

Ever wondered how to make your data say exactly what you want? This tool lets you tweak variables like political affiliation and economic metrics to achieve that elusive "statistical significance". Surely that's just fictional... p-hacking at its finest!

Can we stop with research papers titled '... is all you need'? Seriously, I think we've had enough…

Are PhDs losing their lustre? Why fewer students are enrolling in doctoral degrees High living costs paired with stagnant stipends are being blamed for a drop in PhD enrolments in several countries.

Are PhDs losing their lustre? Why fewer students are enrolling in doctoral degrees

Adaptive XAI Workshop We're thrilled to share the accepted papers for our upcoming workshop in Cagliari. Join us for a day of cutting-edge research exploring adaptive explanations and human-centered approaches to XAI.

📢 Excited to announce the accepted papers for #AXAI2025! From digital twins and emotion recognition to LLM-powered explanations and adaptive interfaces - join us in Cagliari to explore the future of explainable AI. Check out the full list: https://axai.trx.li/accepted-papers/ #IUI2025

The Subtle Art of Designing Physical Controls for Cars Prototyping a concept for simpler, physical climate controls

The Subtle Art of Designing Physical Controls for Cars

Chat is a bad UI pattern for development tools

Life Is More Than an Engineering Problem | Los Angeles Review of Books Julien Crockett speaks with Ted Chiang about the search for a perfect language, the state of AI, and the future direction of technology.

Some of this attitude comes from the fact that the people making AI tools are engineers viewing everything from an engineering perspective, but it's also that, as a culture, we have adopted this way of thinking as the default.

The Irony

"While we encourage people to use AI systems during their role to help them work faster and more effectively, please do not use AI assistants during the application process."

Why do you want to work at Anthropic?
— Anthropic, online job application form

AI is Creating a Generation of Illiterate Programmers A couple of days ago, Cursor went down during the ChatGPT outage. I stared at my terminal facing those red error messages that I hate to see. An AWS error glared back at me. I didn’t want to figure it out without AI’s help. After 12 years of coding, I’d somehow become worse at my own craft. And this isn’t hyperbole—this is the new reality for software developers.

AI is Creating a Generation of Illiterate Programmers

Prototype high fidelity spatial apps.

Funding scientific research… are we doing it well?

Ephemerality in User Interfaces On hypermedia, UI design, and user agency.

Ephemerality in User Interfaces

🔄 Deadline Extension!
AXAI Workshop @ ACM IUI 2025 submissions now due Jan 15th AoE.
Still time to contribute to shaping adaptive & explainable AI interfaces!
Submit at: https://cmt3.research.microsoft.com/AXAI2025/
More info: https://axai.trx.li
#XAI #HCI #IUI2025

An Unreasonable Amount of Time A method for magic.

Sometimes magic is just someone spending more time on something than anyone else might reasonably expect.

🎄 Season's Greetings from the Adaptive XAI Workshop organizers!

As we wrap up 2024, a friendly reminder: Just 2 weeks left to submit your work! Deadline: Jan 13, 2025

Join us at #IUI2025 in Sardinia to shape the future of explainable AI 🔍

Details: https://axai.trx.li

#XAI #HCI #AI

An academic Great Gatsby Curve – How much academic success is inherited? To what extent does academic success follow success? Drawing on a dataset of mentor-mentee relations and their citations, Ye Sun, Fabio Caccioli, Xiancheng Li and Giacomo Livan suggest the dyn…

An academic Great Gatsby Curve – How much academic success is inherited?

Adaptive XAI Workshop @ IUI 2024 Adaptive eXplainable AI - Towards Intelligent Interfaces for Tailored AI Explanations

🚀 Exciting news! The Adaptive XAI Workshop is back for its second edition at #IUI2025 in Cagliari, Sardinia 🇮🇹.

📅 March 24, 2025
📝 Submissions open until Jan 13, 2025

Focusing on tailored AI explanations and intelligent interfaces. Join us! 🔍
👉 axai.trx.li

#AI #XAI #HumanCentricAI #HCI

Keep calm and carry on, with or without AI: What to do if your profession or organization urgently needs to adopt AI

https://iis.seas.harvard.edu/keep-calm-and-carry-on/

Stop using generative AI as a search engine Sometimes ChatGPT spits out something original — and wrong, like fake presidential pardons.

Couldn't have said it better!

Stop using generative AI as a search engine

https://www.theverge.com/2024/12/5/24313222/chatgpt-pardon-biden-bush-esquire

A Voight-Kampff test from Blade Runner.

Grading essays in 2023 AD: