#AI Limitations

Articles (15)

• 🇺🇸 Apple’s AI Playlist Playground is bad at music — 26/03/2026 [USA]

  Apple Music: "What do you want to hear?" Me: "Atmospheric instrumental black metal to write to." Apple Music: "Here's three metal songs with vocals, a field recording, an ambient electronic track, and...

  Related: #Music Curation
• 🇺🇸 Stop asking AI chatbots about your workplace problems—a close colleague is a better bet, psychologist says — 23/03/2026 [USA]

  AI isn't best suited for workplace questions about difficult conversations or social activities, so turn to a close colleague first, says a psychologist.

  Related: #Workplace Advice
• 🇺🇸 Cognitive Mismatch in Multimodal Large Language Models for Discrete Symbol Understanding — 20/03/2026 [USA]

  arXiv:2603.18472v1 Announce Type: new Abstract: While Multimodal Large Language Models (MLLMs) have achieved remarkable success in interpreting natural scenes, their ability to process discrete symbo...

  Related: #Symbolic Reasoning
• 🇺🇸 Large Reasoning Models Struggle to Transfer Parametric Knowledge Across Scripts — 19/03/2026 [USA]

  arXiv:2603.17070v1 Announce Type: cross Abstract: In this work, we analyze shortcomings in cross-lingual knowledge transfer in large, modern reasoning LLMs. We demonstrate that the perceived gap in k...

  Related: #Knowledge Transfer
• 🇺🇸 InPhyRe Discovers: Large Multimodal Models Struggle in Inductive Physical Reasoning — 19/03/2026 [USA]

  arXiv:2509.12263v2 Announce Type: replace Abstract: Large multimodal models (LMMs) encode physical laws observed during training, such as momentum conservation, as parametric knowledge. It allows LMM...

  Related: #Physical Reasoning
• 🇺🇸 Large language models show fragile cognitive reasoning about human emotions — 16/03/2026 [USA]

  arXiv:2508.05880v2 Announce Type: replace-cross Abstract: Affective computing seeks to support the holistic development of artificial intelligence by enabling machines to engage with human emotion. R...

  Related: #Emotional Reasoning
• 🇺🇸 AI Knows What's Wrong But Cannot Fix It: Helicoid Dynamics in Frontier LLMs Under High-Stakes Decisions — 13/03/2026 [USA]

  arXiv:2603.11559v1 Announce Type: new Abstract: Large language models perform reliably when their outputs can be checked: solving equations, writing code, retrieving facts. They perform differently w...

  Related: #Decision-Making
• 🇺🇸 Compression Favors Consistency, Not Truth: When and Why Language Models Prefer Correct Information — 13/03/2026 [USA]

  arXiv:2603.11749v1 Announce Type: cross Abstract: Why do language models sometimes prefer correct statements even when trained on mixed-quality data? We introduce the Compression--Consistency Princip...

  Related: #Model Training
• 🇺🇸 AI layoff reversal: Companies rehire for customer roles they eliminated — 10/03/2026 [USA]

  Complaints from frustrated customers have prompted e-commerce and financial technology companies to quietly rehire some content writers, software engineers and customer service workers they had replac...

  Related: #Employment Trends
• 🇺🇸 Vision Language Models Cannot Reason About Physical Transformation — 10/03/2026 [USA]

  arXiv:2603.07109v1 Announce Type: new Abstract: Understanding physical transformations is fundamental for reasoning in dynamic environments. While Vision Language Models (VLMs) show promise in embodi...

  Related: #Physical Reasoning
• 🇺🇸 Better Eyes, Better Thoughts: Why Vision Chain-of-Thought Fails in Medicine — 10/03/2026 [USA]

  arXiv:2603.06665v1 Announce Type: cross Abstract: Large vision-language models (VLMs) often benefit from chain-of-thought (CoT) prompting in general domains, yet its efficacy in medical vision-langua...

  Related: #Medical Diagnostics
• 🇺🇸 X says you can block Grok from editing your photos — 09/03/2026 [USA]

  X has introduced a new feature that makes it slightly harder for other users to manipulate your uploaded images with the Grok chatbot. As reported by Social Media Today and verified by The Verge, a ne...

  Related: #Social Media
• 🇺🇸 Lost in Stories: Consistency Bugs in Long Story Generation by LLMs — 09/03/2026 [USA]

  arXiv:2603.05890v1 Announce Type: cross Abstract: What happens when a storyteller forgets its own story? Large Language Models (LLMs) can now generate narratives spanning tens of thousands of words, ...

  Related: #Narrative Consistency
• 🇬🇧 Current and former Block workers say AI can’t do their jobs after Jack Dorsey’s mass layoffs: ‘You can’t really AI that’ — 08/03/2026 [United Kingdom]

  <p>The CEO said he cut the company’s workforce by 4,000 people – almost in half – because of gains in AI productivity</p><p>Mark remembers the first time he wondered whether he was teaching Block’s AI...

  Related: #Workforce Impact
• 🇺🇸 The AI Research Assistant: Promise, Peril, and a Proof of Concept — 27/02/2026 [USA]

  arXiv:2602.22842v1 Announce Type: new Abstract: Can artificial intelligence truly contribute to creative mathematical research, or does it merely automate routine calculations while introducing risks...

  Related: #Human-AI Collaboration, #Mathematical Research, #Scientific Discovery