Furthermore, as the method of reporting on perceived and induced affect may influence the construction of an affect (e.g., facilitating categorical perception), we also compare two different affect representations.
makes an explicit connection between a music theory concept and congition
This research follows a constructionist approach to musical affect (Cespedes-Guevara & Eerola, 2018). That is, although we are interested in the \'bottom-up\' influence of certain musical features on musical affect, we believe these cannot be adequately evaluated without considering the \'top-down\' effects of context and individual differences that are present when affects are constructed. The perception or induction of affect does not merely arise in response to a stimulus but is also formed in relation to the individual and the context.
makes an explicit connection between a music theory concept and congition
Cognitive surrenderA paper that came out this year asked: if you’re working with AI a lot, and you’re using it as a machine to answer all of your questions, what happens with System 1 and System 2?
Cognitive surrender: what happens to System 1 and System 2 if you offload to AI to get any answers? (Is this diff from other cognitive tools, like writing and Plato's rejection of it?)
The paper is https://doi.org/10.31234/osf.io/yk25n_v1 and it posits AI offloading as System 3. That is an interesting perspective. Thinking—Fast, Slow, and Artificial: How AI is Reshaping Human Reasoning and the Rise of Cognitive Surrender by Shaw and Nave, 2026. Thinking—Fast, Slow, and Artificial: How AI is Reshaping Human Reasoning and the Rise of Cognitive Surrender in Zotero
2. Validator Another basic role for AI is validating your understanding. To do this, you ask it to review your notes for errors or gaps, do basic fact checking, or critique your reasoning. Again, you can do this via the chat interface, but I also experimented with passing my notes in Obsidian using the Copilot plugin and in Emacs using gptel. Example: After reading The Epic of Gilgamesh, I wrote a note in Obsidian summarizing its plot. When I asked ChatGPT to critique my summary, it pointed out that I’d given the central character a redemption arc that isn’t present in the text. I’m so accustomed to the standard hero’s journey, that I projected it onto the book — and an LLM helped me correct this ‘hallucination.’ Suggested prompt: Here are my notes on [WORK]. What important ideas did I miss or underemphasize? Don’t rewrite my notes — just flag the gaps.
Role 2 validator of one's understanding, also seen as basic. Might be a good complement to e.g. turning some of my notes into [[Anki]] card decks or combine in another way w spaced repetition. [[Spaced repetition 20201012201559]] [[Connecting my PKM to Anki]]
[[Jorge Arango p]] talk at PKM Summit 2026 robots in the garden, a perspective on PKM and the use of gen AI in it
list of scenario's in which AI agents will a) work against you b) be used against you at scale.
For the record, my posts aren’t written or conceived with an LLM, although I know an increasing number of people who use one to write a first draft and then edit. I’m not a fan. The whole point of the web — its beauty — is that it’s unrelentingly human and diverse.
A good case for disfavoring the use of AI/LLMs to write first drafts of blog posts. Implicit I believe is a distinction between using external tools to edit/proofread a human-written draft vs editing/proofreading a machine draft (granting I do not use these tools for either). Related to points I raised in Re; On AI in response to: A Positive Technologist Identity (2/4).
Given the high prevalence of such sounds in everyday life, having misophonia can have large negative effects on one's functioning in personal, academic, and work environments.
any sentences referring to misophonia verbatim
Although there are many idiosyncrasies in what may trigger a person with misophonia, the most common triggers are created by other humans, such as the sound of someone chewing, clearing their throat, tapping their foot, or typing on a keyboard.
any sentences referring to misophonia verbatim
Misophonia is a psychological disorder that is characterized by severe aversive responses to specific environmental sounds (i.e., triggers).
any sentences referring to misophonia verbatim
This indicates that misophonia is not a purely auditory processing disorder but is also influenced by a top-down process of source identification.
any sentences referring to misophonia verbatim
an fMRI study found that people with misophonia show increased response in the anterior insular cortex (AIC) in response to misophonic sounds, compared to control participants and other unpleasant or neutral sounds (Kumar et al., 2017).
any sentences referring to misophonia verbatim
Both the subjective judgment of aversiveness and the physiological measure of skin conductance response (SCR) increase when people with misophonia are presented with triggers (Edelstein et al., 2013).
any sentences referring to misophonia verbatim
The disorder is not yet recognized by the Diagnostic and Statistical Manual − 5th version (DSM-5; American Psychiatric Association, 2013), but there has been an increasing amount of research on the characterization and treatment of misophonia (Vitoratou et al., 2021; see also Brout et al., 2018, for a review).
any sentences referring to misophonia verbatim
Composers and music researchers had previously analyzed and annotated 65 movements from the Classical, Romantic, and early Modern repertoire in terms of the Taxonomy of Orchestral Grouping Effects (McAdams et al., 2022).
please find any claims that depend on citations referring to works by any of the present authors
In a study by McAdams and Goodchild (2018), orchestral simulations created with OrchSim were compared perceptually to commercial recordings and were shown to be of high quality.
please find any claims that depend on citations referring to works by any of the present authors
These results confirm with orchestral excerpts the findings of studies on isolated tones with dyads or triads of instruments in which the presence of impulsive instruments reduces the perception of blend (Lembke et al., 2019; Reuter, 1996; Tardieu & McAdams, 2012).
please find any claims that depend on citations referring to works by any of the present authors
This provides additional support for its significant contribution to blend in Fischer et al. (2021).
please find any claims that depend on citations referring to works by any of the present authors
Lembke et al. (2017) demonstrated that combinations of sustained and impulsive instruments blend less well.
please find any claims that depend on citations referring to works by any of the present authors
structuring by affecting sequential grouping through the segregation of auditory streams played by different instruments and segmental grouping through timbral contrasts (McAdams et al., 2022).
please find any claims that depend on citations referring to works by any of the present authors
Several other spectral and spectrotemporal descriptors were found to play a role in blend perception in orchestral works by Fischer et al. (2021). These include spectral flatness and spectral crest (different measures of the degree to which the spectrum is denser or has more emergence of spectral components), and spectral variation (the degree of variation of the spectral shape over time).
please find any claims that depend on citations referring to works by any of the present authors
Fischer et al. (2021) studied the blends of multi-instrument streams in the context of orchestral stream segregation in predominantly Romantic orchestral excerpts. They found that within-family instrument combinations blended better than between-family combinations. They demonstrated the role played by overlap in timbre correlates of spectral flatness (a measure of the tonalness/noisiness or density of the spectrum), spectral skewness (related to the shape of the spectral envelope), and spectral variation (evolution of the spectral envelope over time), as well as cues derived from the scores such as onset synchrony and the consonance of concurrent pitch relations.
please find any claims that depend on citations referring to works by any of the present authors
McAdams et al. (2022) distinguish two subcategories of these two types of blend: stable and transforming.
please find any claims that depend on citations referring to works by any of the present authors
Lembke, S.-A., Parker, K., Narmour, E., & McAdams, S. (2019). Acoustical correlates of perceptual blend in timbre dyads and triads. Musicae Scientiae, 23(2), 250–274.
please find any claims that depend on citations referring to works by any of the present authors
Several other spectral and spectrotemporal descriptors were found to play a role in blend perception in orchestral works by Fischer et al. (2021).
please find any claims that depend on citations referring to works by any of the present authors
Lembke and McAdams (2015) found that the degree of spectral overlap between constituent sounds enhanced blend perception.
please find any claims that depend on citations referring to works by any of the present authors
Tardieu and McAdams (2012) extended this work with combinations of unison sustained and impulsive instruments (including pitched percussion and string pizzicati).
please find any claims that depend on citations referring to works by any of the present authors
McAdams et al. (2022) introduce other notions related to blend as well.
please find any claims that depend on citations referring to works by any of the present authors
This in turn creates an effect of perceptual unity (McAdams, 1989).
please find any claims that depend on citations referring to works by any of the present authors
Four important concurrent grouping cues predict the perceptual fusion of sound events (McAdams, 1984): onset synchrony, harmonicity, coherent frequency behavior, and common spatial origin.
please find any claims that depend on citations referring to works by any of the present authors
To this, McAdams et al. (2022) have added segmental grouping (chunking into perceptual units).
please find any claims that depend on citations referring to works by any of the present authors
When the sudden drop to a pianissimo occurred towards the ending of the piece, the perceived arousal responses of CHM and WM dropped slightly but rose again immediately to end on a high arousal. These two groups of listeners appear to have anticipated a return to a loud and majestic close and therefore kept their arousal responses higher than those of the NM.
please highlight anything related to music performance practice
CHM, who are more experienced with the instruments and compositional techniques used in Chinese orchestral music, might have had an idea of which features figure more prominently in the communication of particular intentions, and therefore would have more information available for their judgments.
please highlight anything related to music performance practice
The perception of affective intentions in music is influenced by the degree of familiarity listeners have with a musical tradition, the content implicated in the music, and the complex sonic environment created by the composer's creation and the musicians' interpretation.
please highlight anything related to music performance practice
The version that participants heard was a premier of the work by the Taipei Chinese orchestra.
please highlight anything related to music performance practice
The communication of emotions or affect takes place when listeners perceive emotional meaning that is expressed by performers in music (Juslin, 2013a, 2013b).
please highlight anything related to music performance practice
An understanding of a tonal schema with its associations to happiness and sadness has been consistently found to influence listeners who have grown up in a culture of Western music.
please highlight anything related to music theory
Its mirmode function estimates the modal strength of the music in terms of ''majorness'' and ''minorness.''
please highlight anything related to music theory
The perception of consonance also plays an important role in the music listening process—combinations of tones that are consonant are perceived as more positively valenced than dissonant ones (Harrison & Pearce, 2020).
please highlight anything related to music theory
Musical structures such as pitch relations are perceptually salient and provide important information for listeners (e.g., Gabrielsson & Lindstrom, 2010; Krumhansl, 1998; Krumhansl & Kessler, 1982).
please highlight anything related to music theory
Iqa' (plural iqa'at) is used to describe a rhythmic cycle. Iqa'at are made up of two different basic building blocks, the dum and tak, onomatopoeias derived from the sound produced on membranophones such as the darabuka.
please highlight anything related to music theory
H5. Being more culturally bound, musical cues that are learned, such as modal structures, metrical relations, and so on, will exert a greater influence on listeners' perceived valence ratings than on their arousal ratings.
please highlight anything related to music theory
This is a simple PDF file. Fun fun fun.
testing something new today
using SemanticCommit, we recorded all instances of edits, check for conflicts, make change, local, and global resolution actions using telemetry.
sentences describing methods the authors used; one sentence at a time
A sub-task was considered a failure if the participant was unable to complete it within the time limit.
sentences describing methods the authors used; one sentence at a time
Finally, we conducted an informal interview about their experience.
sentences describing methods the authors used; one sentence at a time
After both tasks were completed, participants filled out a final survey to compare the two conditions.
sentences describing methods the authors used; one sentence at a time
We chose GPT-4o for performance and latency reasons, as it performed optimally against our evals.
sentences describing methods the authors used; one sentence at a time
We also ran evaluations of model latency and classification performance under varying false positive rates for the following LLMs by OpenAI: GPT-4o, GPT-4o-mini, and o3-mini.
sentences describing methods the authors used; one sentence at a time
For each task, participants were tasked with integrating three new pieces of information into the memory, one at a time ("sub-tasks").
sentences describing methods the authors used; one sentence at a time
We ensured each list was 30 items long as our pilot studies suggested this was long enough that manual detection starts to become unwieldy (users need to scroll up and down the document), but short enough that participants could become familiar in a short period.
sentences describing methods the authors used; one sentence at a time
We adapted two intent specifications from our evals: Mars Game Design Document and Financial Advice AI Agent Memory, as these tasks mapped to the two paradigmatic types covered in Sections 2 and 2.1 (design documents, and AI memory of the user).
sentences describing methods the authors used; one sentence at a time
We recruited 12 participants (7 female, 5 male) through the mailing lists of two research universities and one multinational technology company.
sentences describing methods the authors used; one sentence at a time
We chose OpenAI's ChatGPT Canvas as a baseline for five reasons: (i) it is a popular, commercially available tool, hence it is likely familiar to users; (ii) it provides a document editing view, where users can select text and ask GPT to rewrite it, or chat with an AI to make global edits; (iii) it employs a similar class of model (GPT-4o); (iv) it supports similar editing features as SemanticCommit like inline text selection, conflict highlighting, and a diff view, while adding free-form editing; and (v) similar interfaces like Anthropic Artifacts tended to rewrite the specification entirely, and did not offer Canvas's "diff" view to allow for a fair comparison.
sentences describing methods the authors used; one sentence at a time
With participant consent, we recorded audio and screen-casts, and participants were encouraged to think aloud.
sentences describing methods the authors used; one sentence at a time
Four coauthors created the evals, and two coauthors manually double-checked all conflicts, a process that took several days.
sentences describing methods the authors used; one sentence at a time
We ran one pilot study with five users of our card-based interface, and a second with four users of a revised interface.
sentences describing methods the authors used; one sentence at a time
Our explorations went through substantial iterations and prompt prototyping over a period of eight months, evolving in response to two pilot studies and progressing from a card-based interface to a list of texts.
sentences describing methods the authors used; one sentence at a time
We iterated on prompts using ChainForge [5] by setting up an evaluation pipeline against our datasets, which allowed us to observe the effects of prompt changes and model choices.
sentences describing methods the authors used; one sentence at a time
To measure statistical significance, we used Mann–Whitney–Wilcoxon tests and report the p-values.
sentences describing methods the authors used; one sentence at a time
For qualitative analysis, the first author performed open coding on participant responses and audio transcripts to identify themes, which were used to interpret the qualitative results.
sentences describing methods the authors used; one sentence at a time
In the post-task surveys, we collected self-reported NASA Task Load Index (TLX) scores, Likert-scale ratings for ease of use, and responses on how well the AI helped participants identify, understand, and resolve semantic conflicts.
sentences describing methods the authors used; one sentence at a time
Each condition had a time limit of 15 minutes, after which the participant completed a post-task survey.
sentences describing methods the authors used; one sentence at a time
Before each task, participants received a tutorial on the assigned tool and were given five minutes to explore it using a test document.
sentences describing methods the authors used; one sentence at a time
Both the order of task assignment and tool assignment were counterbalanced and randomly assigned.
sentences describing methods the authors used; one sentence at a time
We conducted a controlled within-subjects study with mixed methods, comparing SemanticCommit with a baseline interface.
sentences describing methods the authors used; one sentence at a time
We run end-to-end on our four eval datasets using GPT-4o and GPT-4o-mini and report the mean ± stddev for accuracy, precision, recall, and F1 scores for the three approaches in Figure 5.
sentences describing methods the authors used; one sentence at a time
We compare our end-to-end system against two simpler methods: (i) DropAllDocs, which adds all documents to the context for conflict classification; and (ii) InkSync [56] which generates a JSON list of string-replace operations.
sentences describing methods the authors used; one sentence at a time
In order to minimize relevance assessment issues, we apply a PageRank-based relevance ranking over the KG, akin to HippoRAG [36].
sentences describing methods the authors used; one sentence at a time
We implement the back-end using a knowledge-graph (KG) RAG architecture [36] consisting of two phases: pre-processing and inference.
sentences describing methods the authors used; one sentence at a time
Through a within-subjects study with 12 participants comparing SemanticCommit to a chat-with-document baseline (OpenAI Canvas), we find differences in workflow: half of our participants adopted a workflow of impact analysis when using SemanticCommit, where they would first flag conflicts without AI revisions then resolve conflicts locally, despite having access to a global revision feature.
sentences describing methods the authors used; one sentence at a time
We implement the back-end using a knowledge-graph (KG) RAG architecture [36] consisting of two phases: pre-processing and inference.
sentences describing methods the authors used; one sentence at a time
In order to minimize relevance assessment issues, we apply a PageRank-based relevance ranking over the KG, akin to HippoRAG [36].
sentences describing methods the authors used; one sentence at a time
We compare our end-to-end system against two simpler methods: (i) DropAllDocs, which adds all documents to the context for conflict classification; and (ii) InkSync [56] which generates a JSON list of string-replace operations.
sentences describing methods the authors used; one sentence at a time
We run end-to-end on our four eval datasets using GPT-4o and GPT-4o-mini and report the mean ± stddev for accuracy, precision, recall, and F1 scores for the three approaches in Figure 5.
sentences describing methods the authors used; one sentence at a time
We conducted a controlled within-subjects study with mixed methods, comparing SemanticCommit with a baseline interface.
sentences describing methods the authors used; one sentence at a time
Both the order of task assignment and tool assignment were counterbalanced and randomly assigned.
sentences describing methods the authors used; one sentence at a time
Before each task, participants received a tutorial on the assigned tool and were given five minutes to explore it using a test document.
sentences describing methods the authors used; one sentence at a time
Each condition had a time limit of 15 minutes, after which the participant completed a post-task survey.
sentences describing methods the authors used; one sentence at a time
In the post-task surveys, we collected self-reported NASA Task Load Index (TLX) scores, Likert-scale ratings for ease of use, and responses on how well the AI helped participants identify, understand, and resolve semantic conflicts.
sentences describing methods the authors used; one sentence at a time
For qualitative analysis, the first author performed open coding on participant responses and audio transcripts to identify themes, which were used to interpret the qualitative results.
sentences describing methods the authors used; one sentence at a time
To measure statistical significance, we used Mann–Whitney–Wilcoxon tests and report the p-values.
sentences describing methods the authors used; one sentence at a time
We iterated on prompts using ChainForge [5] by setting up an evaluation pipeline against our datasets, which allowed us to observe the effects of prompt changes and model choices.
sentences describing methods the authors used; one sentence at a time
Our explorations went through substantial iterations and prompt prototyping over a period of eight months, evolving in response to two pilot studies and progressing from a card-based interface to a list of texts.
sentences describing methods the authors used; one sentence at a time
We ran one pilot study with five users of our card-based interface, and a second with four users of a revised interface.
sentences describing methods the authors used; one sentence at a time
Four coauthors created the evals, and two coauthors manually double-checked all conflicts, a process that took several days.
sentences describing methods the authors used; one sentence at a time
With participant consent, we recorded audio and screen-casts, and participants were encouraged to think aloud.
sentences describing methods the authors used; one sentence at a time
These semantic conflicts require dedicated support to detect, visualize, and resolve. Semantic conflict resolution interfaces must go beyond visualizing what changes were made, to what changes could be made, where they should be made, and what the effects might be. This resembles feedforward: affordances that help the user foresee the impact of an action [67, 93].
sentences describing connections to theory; one sentence at a time
Today with LLMs, we are less limited by this constraint, and solutions to the problem of human-machine communication might be better found in cybernetics theory [9] than static formalism.
sentences describing connections to theory; one sentence at a time
This reflects the principle of feedforward [67, 93] in communication theory—"a needed prescription or plan for a feedback, to which the actual feedback may or may not confirm" [79]—where a communicator provides "the context of what one was planning to talk about" [64, p. 179-80] in order to "pre-test the impact of [its output]" on the listener [34, p. 65].
sentences describing connections to theory; one sentence at a time
One illustration is using menus, instructions, graphical user interfaces (GUI), or virtual reality to access any particular computer's features.
sentence about GUIs
An appealing alternative to conventional text-based interfaces through graphical user interfaces is the direct use of hands as an input device to provide natural human-computer interaction.
sentence about GUIs
A more thorough description of the current tools and techniques for interacting with computers as well as recent developments in the subject is provided in the next section.
sentence about GUIs
The evolving multi-modal and Graphical user interfaces (GUI) enable humans to interact with embodied character agents in a way that is not possible with other interface paradigms.
sentence about GUIs
The widely used graphical user interfaces (GUI) of today are found in desktop applications, internet browsers, mobile computers, and computer kiosks.
sentence about GUIs
and the narrowing of the medullary cavity of tubular bones (C) (diaphyseal stenosis).
Subject ID: Individual_2 Phenotype: Narrowing of the medullary cavity of tubular bones
In this work, we introduce a new paradigm for exploring a large corpus of small documents by identifying roles at the phrasal and sentence levels, then slice on, reify, group, and/or align the text itself on those roles, with sentences left intact.
please find me the main contributions of this paper
AbstractExplorer instantiates new minimally lossy SMT-informed techniques for skimming, reading, and reasoning about a corpus of similarly structured short documents: phrase-level role classification that drives sentence ordering, highlighting, and spatial alignment.
please find me the main contributions of this paper
AbstractExplorer has a unique combination of LLM-powered (1) faceted comparative close reading with (2) role highlighting enhanced by (3) structure-based ordering and (4) alignment. An ablation study (N=24) validated that these features work best together. A summative study (N=16) describes how these features support users in familiarizing themselves with a corpus of paper abstracts from a single large conference with over 1000 papers.
please find me the main contributions of this paper
We contribute: • Novel SMT theory-informed text analysis and rendering techniques for enabling cross-document skimming and comparative close reading at scale • AbstractExplorer, which instantiates these techniques for familiarizing oneself with a corpus of ∼1000 CHI paper abstracts. • Three studies informing and evalutaing the benefits, challenges, and interactions between these techniques.
please find me the main contributions of this paper
for - AI - Bayesian teaching paper - Google
for - AI - biological neural networks - AI - mortal computing
Ai and taxation
However, in a narrow set of cases, we believe AI can undermine, rather than defend, democratic values. Some uses are also simply outside the bounds of what today’s technology can safely and reliably do. Two such use cases have never been included in our contracts with the Department of War, and we believe they should not be included now:
Humans as API, bc [[Alles is een API 20260309095254]]
(via [[Stephen Downes p]]
the 'am dash', as opposed to the em dash. A human only glyph that AI 'can't use'
I tnink there I am dash
The real annoying thing about Opus 4.6/Codex 5.3 is that it’s impossible to publicly say “Opus 4.5 (and the models that came after it) are an order of magnitude better than coding LLMs released just months before it” without sounding like an AI hype booster clickbaiting, but it’s the counterintuitive truth to my personal frustration
A generative AI like ChatGPTData Analyst can take on the role of the evaluation soft-ware. It is expected that this manner of use will make thestudents' work easier, as less emphasis needs to be placedon the programming itself. Instead, teachers can incorpo-rate exercises that encourage students to code more effi-ciently and accurately with the assistance of AI. Thisshifts the focus from finding the right command or func-tion to examining and understanding the data moreclosely. As a consequence, students are better enabled tointerpret the results of statistical evaluation software cor-rectly, thus fulfilling goal 8 of the GAISE report.
rhetoric: Schwarz uses a statement of transition to contrast the old education model (rote memorization of commands) with a new required model (critical examination).
inference: This supports the argument that education and labor must start to pivot away from the "Generalist" process-oriented tasks. If the machine assistants handle the 'How' (the commands and functions), then the human must focus more on the 'Why' and the 'what does it mean (understanding/wisdom)'. This helps to validate the work of the assistants and helps to make it useful and valuable in the real world.
statistical knowledge is still required in order toformulate the correct prompts and to ensure that the AIdoes not leave out any step of the analysis.
rhetoric: author presents a prescriptive claim that AI needs humans with competent knowledge (in this case, statistics) to create prompts and ensure that the AI does not leave out any steps of the analysis. He positions domain knowledge not as a tool for using AI for statistical analysis, but a prerequisite for management of the AI and auditing the output.
inference: In addition to policing and correcting the AI outputs, the deep domain knowledge is what allows the AI to do complex data analysis without mistakes, hallucinated results, or mathematically false outcomes. This is basically the job description of a human with "Augmented Human Wisdom". The human's value is no longer in doing math, but in possessing the vertical expertise (flesh/wisdom) to know exact what math needs to be done and ultimately auditing the assistant machine's work.
ChatGPT Data Analyst clearly produced a false resulthere, precisely because the application assumptions for theANOVA were not checked.
rhetoric: Schwarz employs cause-and-effect reasoning here based on empirical testing. He links a specific technical failure (not checking assumptions) to a definitive unwanted outcome (a false result).
inference: the "Data Analyst" function of ChatGPT hallucinated a result during the use of it's core function! This is the best evidence so far of the 'Crisis of Truth' and the dangers of the 'Headless Automatons' in my essay. If a generalist with no deep knowledge uses AI, they are at great risk of blindly accepting mathematically false conclusions. Synthetic syntax without competent human validation is a liability.
The results show that generative AI canfacilitate data analysis for individuals with minimal knowledge of statistics,mainly by generating appropriate code, but only partly by following standardprocedures.
rhetoric: author uses comparative, objective statement (logos) to establish the main boundary of the technology's capability/capacity -- it excels at technical generation (things like coding) but fails at standard procedures (methodological adherence to SOPs).
inference: the proves the 'Raising the Floor' concept. AI completely automates the entry-level syntax (the "Word"), meaning that the Generalist coder is obsolete! However, because it fails at standard procedures, it requires a human architect to guide it to outputs that are valuable in the real world.
PWA have language deficits that require bespoke AAC supports. These supports may beenhanced by LLMs in software systems that use spoken user input to provide relevantsuggestions that have grammatical and speech production support.
rhetoric: concluding statement. this positions the LLM as an 'enhancement' to physical human limitation, rather than a replacement of the human subject.
inference: This helps to validate the 'Augmented Human Wisdom' model. The future of AI is NOT replacing humans, but AI acting as a high-powered syntax engine that is strictly guided by human needs and human intent. The AI does not have 'agency', as it is a software tool that helps the human to execute their visions.
Perseverations that are input into the system are essentially mag-nified by the system’s suggested sentences,
rhetoric: authors explain an unintended consequence of using the AI tool: it scales the errors or the emptiness of the human prompt.
inference: this is an excellent metaphor for the 'manager fallacy'. If the human user in incompetent (or provides empty or incomplete input), the AI does not magically create wisdom -- it just amplifies the user's incompetence in a a highly articulate synthetic thought.
Participant 2 stated the age of her daughters (“Name1 is 18, Name2 is21”), Aphasia-GPT transformed it as “Name1 is 18 and 21”, which is an impossible, butrelated, hallucination
rhetoric: researchers use a specific, clinical observation of an error to demonstrate the model's inability to comprehend logical reality despite the human relaying a perfectly structured sentence.
inference: this shows that AI is amoral and lacks the lived experience necessary to make logical judgments that work in the real world. It can format a sentence beautifully, but it does not/will not always understand that a single human cannot be two ages at once. This is why it is very important/necessary for the "flesh" to text the output against reality
Aphasia-GPT is a real-time, AI-enabled web app designed to expand the words providedby a user into complete sentences as suggestions for a user to select.
rhetoric: authors provide a definition of their creation (Aphasia-GPT) to describe it's mechanism: taking a fragmented input and expanding it into a fully structured, complete output.
inference: this is the embodiment of Harari's primary metanym of the word v flesh (syntax v human). in this example, Aphasia-GPT provides the words (syntax) to the fleshy human that struggles with those words, while also relying on the human to spark the intent of the communication. The human is using AI to communicate with words, because the words are very difficult for the human.
nearly half of the workers who received sloppy AI-generated material from coworkers saw those coworkers as less creative, capable and reliable, while smaller amounts saw them as less trustworthy and intelligent.
AI Slop generated by your own hands changes he way people view you at work!
The cost of the time that it takes fix "workslop" could add up too, with a $186 monthly cost per employee on average, according to a survey of desk workers by BetterUp in partnership with the Stanford Social Media Lab. Forty percent of the workers surveyed said they received "workslop" in the last month and that it took an average of two hours to resolve each incident.
$186/per employee/per month!
10 employees = ($22,320) 25 employees = ($55,800) 50 employees = ($111,600) 100 employees = ($223,200) 250 employees = ($558,000) 500 employees = ($1,116,000) 1000 employees = ($2,232,000)
“Younger workers aren’t necessarily more careless, but they’re often using AI more frequently and earlier in their workflows," Dennison said. "There is also a training gap. Organizations often assume younger employees intuitively understand AI, yet provide little guidance on verification, risk, or appropriate use cases. As a result, AI may be treated as an answer engine rather than a support tool."
this is another great quote, which helps to establish how orgs treat younger generations, and how they tend to overtrust their understanding of AI.
About 70% of managers say their direct reports have made AI-related mistakes in the past year
Good stat from a survey of 1100 managers at resume.com. 70% of managers say that their employees have submitted AI generated errors inside their work in the last year.
58% said direct reports submitted work that contained factual inaccuracies generated by AI tools, while fewer reported that AI failed to account for critical contextual factors. Other issues cited include low-quality content, poor recommendations and inappropriate messaging.
from reporting managers, 58% of them said that employees were submitting work that contained factual inaccuracies in the work that was generated by AI, and that fewer of them reported that AI failed to account for "critical contextual factors", implying that the writing was generic and not directly applicable to the context that the writing was written in. Other issues were: low quality content, poor recommendations and inappropriate messaging.
59% of managers saying that they had to invest additional time to correct or redo work created by AI. Similarly, 53% said their direct reports had to take on extra work, while 45% said they had to bring in co-workers to help fix the mistake.
Extra time and money spent to repair errors made by AI but not caught by the human in the middle. 59% is almost 2/3 (closer to 3/5) needed to correct or redo the work created by AI without a human auditing it. 53% claim extra work is needed to repair the AI mistakes, and 45% also needed to bring in a (perhaps more senior) co-worker to help fix the mistake. I can imagine workers needing to work on a mistake the hits production code, and all of the thousands (or more) mistakes that would need to be later repaired and rolled back. very expensive and costly.
they damaged client relationships.
missed deadlines are tough, but damaging client relationships is a sin!
While 18% of managers said they did not suffer any financial losses from the mistakes, and 20% said those losses were less than $1,000, a significant number reported bigger losses. Twelve percent said those losses were more than $25,000, while 11% said between $10,000 and $24,999. Another 27% placed the value of those losses above $1,000 but below $10,000.
great stats for the cost of using AI without human auditing.
34% said Gen Z followed by 26% who blamed millennials, with 18% citing Gen X and just 9% saying it was baby boomers.
while this affect all generations, this is clearly a young people problem! This problem is here to stay
“AI is reliable when used as an assistant, not a decision-maker," Dennison said. "Without human judgment and clear processes, speed becomes a risk, and efficiency gains can turn into costly mistakes,”
great quote. directly mentions my concept of requiring human judgement, and how not having a human in the loop can make work move faster, but can also lead to very costly mistakes.
“Employees treat AI outputs as finished work rather than as a starting point. Current AI tools are very good at generating fluent content, but they don’t understand context, business nuance, risk, or consequences. That gap shows up in factual errors, missing constraints, poor judgment calls, and tone misalignment.”
another great quote -- ties into the abdicating human agency to a robot, and the full quote even illustrates the dangers of doing so.
AI fatigue is real and nobody talks about it
[[AI Contintent Action Plan]]
[[The Means of Production by Maximilian Kasy]] overview in slides by author https://maxkasy.github.io/home/files/slides/meansofprediction_slides_long.pdf
On ai, productivity and shorter work weeks and why that will not happen
The scenarios Wooldridge imagines include a deadly software update for self-driving cars, an AI-powered hack that grounds global airlines, or a Barings bank-style collapse of a major company, triggered by AI doing something stupid. “These are very, very plausible scenarios,” he said. “There are all sorts of ways AI could very publicly go wrong.”
Scenario's for a Hindenburg style event: - deadly software update for self driving cars - AI-powered hacking ground global airlines (not sure, if that is clear enough to people, unlike the self driving cars running amok) - Barings-style collapse of a major company triggered by AI (if it's a tech company, it may be less shock, more ridicule, but still)
“It’s the classic technology scenario,” he said. “You’ve got a technology that’s very, very promising, but not as rigorously tested as you would like it to be, and the commercial pressure behind it is unbearable.”
true for AI, but wasn't the case for Hindenburg I'd say.
The race to get artificial intelligence to market has raised the risk of a Hindenburg-style disaster that shatters global confidence in the technology, a leading researcher has warned.Michael Wooldridge, a professor of AI at Oxford University, said the danger arose from the immense commercial pressures that technology firms were under to release new AI tools, with companies desperate to win customers before the products’ capabilities and potential flaws are fully understood.
prediction Michael Wooldridge (Oxford, AI), sees a risk at an 'Hindenburg' event. Shattering the global confidence in AI tech. I"m not sure this analogy entirely fits other than in its potential impact (AI isn't globally trusted, the Hindenburg did not fail bc of the tech itself but bc helium not being allowed to export from the US at the time. Still the Hindenburg did put an end to the entire zeppelin industry yes. No matter the causes.)
OpenClaw, like many other open-source tools, allows users to connect to different AI models via an application programming interface, or API. Within days of OpenClaw’s release, the team revealed that Kimi’s K2.5 had surpassed Claude Opus and became the most used AI model—by token count, meaning it was handling more total text processed across user prompts and model responses.
Wow, I had no idea that Kimi 2.5 had subbed in for Claude Opus so quickly.
OECD catalogue of AI tools en metrics for trustworthy AI.
Low-cost Chinese AI models forge ahead, even in the US, raising the risks of a US AI bubble Nvidia’s latest earnings report reassured some. But Chinese AI models are fast gaining a following around the world, underlining concerns over an ‘AI bubble’ centered on high-investment, high-cost US models.
One of the largest PC suppliers, Dell, was reported to be planning a price hike that could raise hardware costs by hundreds of dollars. Interestingly, for consumers opting for higher memory configurations, this would now require a significant price increase. Here were the price increases that were reported across a variety of products: $130–$230 increase for Dell Pro and Pro Max notebooks and desktops configured with 32 GB of memory $520–$765 increase for systems configured with 128 GB of memory $55–$135 increase for configurations with a 1 TB SSD $66 increase for AI laptops equipped with an NVIDIA RTX PRO 500 Blackwell GPU (6 GB) $530 increase for AI laptops equipped with an NVIDIA RTX PRO 500 Blackwell GPU (24 GB) Similarly, companies like ASUS and Acer were also reported to be bumping up PC pricing to cope with memory shortages, and according to Acer's Chairman, Jason Chen, the BoM (Bill of Materials) for several products within Acer's portfolio has risen dramatically, leaving no choice but to increase prices to ensure consistent supply. Small-scale manufacturers like Framework are also looking to increase the cost of upgrading RAM on existing configurations, indicating a widespread "price hike" wave approaching gamers.
price hikes of DRAM, due to pc laptop manufacturers having trouble in getting enough RAM. Shortages to keep going for 2026, after 2025. AI supply chain gobbling up the rest.
https://web.archive.org/web/20260216112337/https://wccftech.com/western-digital-has-no-more-hdd-capacity-left-out/ Western Digital says consumer sales of HDDs is now just 5% of their sales, everything else going to AI companies. 2026 WD is already sold out. For 2027 and 2028 there is already a few deals in place too.
the humans involved may have simply lost the plot and may not understand what the program is supposed to do, how their intentions were implemented, or how to possibly change it.
key imo. generating code / material, can quickly mean loss of overview (I see how that happens in my use of #algogens if I don't explicitly counteract it), uncertainty about how demands were implemented, and thus what entry points for change there are.
AI infrastructure developers cannot wait five years. In many cases, they cannot wait six months, because waiting six months costs billions of dollars of lost opportunities.
The quick very rough mental maths on a GW of capacity being worth 10billion USD converts to between 1000-1500 USD per megawatt hour of money they think they could be making if they could sell the compute it powered
https://web.archive.org/web/20260215105347/https://simonwillison.net/2026/Feb/15/cognitive-debt/
Simon Willison on cognitive debt (the consequences of vibecoding more or less).
A description of the type of things Activate does w AI agents. Ref'd in [[Paolo’s Weblog It’s still time for free AI love]]
we might move again. The point is that we can. We can because we own our prompts, our skills, our databases, our memory architecture, they all live in our bar. None of it lives inside OpenAI or Anthropic. When we moved, we rewired the model layer and everything else stayed put. That’s the whole trick, really. If you control the pieces that make your agents smart, switching the engine underneath is just plumbing.
Description of how Activate keep their prompts, skills, databases, memory architecture under their own control and within their own environment.
Moving means wiring up another model or models, but the rest is kept as is.
[[Paolo Valdemarin p]] on avoiding lock-in in a single specific AI silo. Bc what is state of the art will shift every few months.
What if I actually did have dirt on me that an AI could leverage? What could it make me do? How many people have open social media accounts, reused usernames, and no idea that AI could connect those dots to find out things no one knows?
AI agents as kompromat collectors
AI Doesn’t Reduce Work—It Intensifies It
I’m going to cure my girlfriend’s brain tumor.
Articles and books by [[Paul Nemitz c]] (als co-authoring w [[Matthias Pfeffer p]]
this reminds me of installation art that is just a prescription / recipe of how to recreate it.
Owning a $5M data center
I miss thinking hard.
AI voor woningbouw
hopelijk niet genAI
Sovereign Minds Leadership and Work in the AI Era Structure, Agency, and Judgment Under Acceleration
[[Sovereign Minds by Nadia Alter]] looks very worthwhile to get. mentions agency in the subtitle.
(via [[Alberto Cottica p]] on masto)
on responsibility, decision making and authority in organisations getting amorph, even before AI. Vgl [[Corporations as Slow AI 20180201210258]]
via [[Alberto Cottica p]] on Masto
The pipeline runs from women as the original “computers” in the 1940s, through the masculinization of computing that pushed women into typing pools and administrative support, through the automation of those roles, to AI assistants today automating what remains: scheduling, reminding, organizing, emotional management.
There is a line from computers in the original sense, to typing pools, admin support, to automation to AI.
“Obedient and obliging machines that pretend to be women are entering our homes, cars and offices,” warned UNESCO’s Director for Gender Equality, Saniye Gülser Corat, in the agency’s landmark 2019 report.
Unesco report, Saniye Gülser Corat (dir for gender equality).
Standard Retrieval-Augmented Generation (RAG) over documents is a good first step, but it fails when faced with complex, cross-domain enterprise questions. It finds text that looks similar, which isn’t the same as finding facts that are related.
criticism of retrieval augmented generatio (RAG): fails in cross domain settings, finds similar text not relations between facts or meaning
openclaw 'ai personal assistant' that you can interact w through your regular chat apps it says
the skill.md for Moltbook.
'a social network for AI agents' wut?
for - essay Claude CEO - risks of AI
steer AI models, and to shaping the character and constitution of these models
for - AI constitution - Deep Humanity?
I do see a path to a slight moderation in AI development that is compatible with a realist view of geopolitics
for - Deep Humanity - AI crisis - spiritual singularity
By the nature of unknown unknowns it is impossible to make an exhaustive list,
for - Johari Window - unknown unknown - AI
AI companies, semiconductor companies, and perhaps downstream application companies generating ~$3T in revenue per year,
for - progress trap - AI - inequality
we cannot ignore the potential for abuse of these technologies by democratic governments themselves.
for - progress trap - AI - democracies becoming autocracies - Hitler - Trump
brainwashing
for - progress trap - AI - propaganda - brainwashing
a true panopticon on a scale that we don’t see today
for - progress trap - AI - Panopticon - Orwellian
Putting these two concerns together leads to the alarming possibility of a global totalitarian dictatorship. Obviously, it should be one of our highest priorities to prevent this outcome.
for - progress trap - AI - arms race amongst dictators! - Trump - Putin - Kim Jong Un
authoritarian governments might use powerful AI to surveil or repress their citizens in ways that would be extremely difficult to reform or overthrow.
for - progress trap - AI - misuse of power - Palantir?
has a small number of specific hard-line prohibitions, and one of them relates to helping with the production of biological (or chemical, or nuclear, or radiological) weapons. But all models can be jailbroken
for - Claude AI - prohibition against developing weapons - can be jailbroken
a genius in everyone’s pocket could remove that barrier, essentially making everyone a PhD virologist who can be walked through the process of designing, synthesizing, and releasing a biological weapon
for - progress trap - AI - technology as an amplifier - technology acts as an amplifier, allowing humans to fly, to move at speeds faster than any known animal, to lift things no living creature can, etc - The danger is ignorance and polarized views combined with extreme self-rightiousness
I believe the only solution is legislation—laws that directly affect the behavior of AI companies,
for - progress trap - AI - legislation required - tell it to Trump!
not all AI companies do this, and the worst ones can still be a danger to everyone even if the best ones have excellent practices.
for - progress trap - AI - rogue AI vendors - like Grok
a clockwork watch may be ticking normally, such that it’s very hard to tell that it is likely to break down next month, but opening up the watch and looking inside can reveal mechanical weaknesses that allow you to figure it out.
for - AI - progress trap - interpretability testing - deception - Is it possible that AI could even change their node behavior as a deceptive move?
similar to how neuroscientists study animal brains by correlating measurement and intervention to external stimuli and behavior
for - AI - testing - interpretability - neuroscience techniques
AI models are grown rather than built
for - AI - training - grown, not built
almost always follows this constitution
for - progress trap - AI - hackers - hack the AI constitution
Constitutional AI,
for - definition - constitutional AI
The concern is that there is some risk (far from a certainty, but some risk) that AI becomes a much more powerful version of such a person, due to getting something wrong about its very complex training process.
for - progress trap - AI - metaphor - psychopath with a nuclear bomb
AI models could develop personalities during training that are (or if they occurred in humans would be described as) psychotic, paranoid, violent, or unstable, and act out, which for very powerful or capable systems could involve exterminating humanity.
for - progress trap - AI - abstraction - progress trap - AI with feelings & AI without feelings - no win? - One major and obvious aspect of current AI LLMs is that they are not only artificial in their intelligence, but also artificial in their lack of real world experiences. They are not embodied (and it would likely be a highly dubious ethical justification for their embodiment as in AI - powered robots) - Once we have the first known AI robot killing a human, it will be an indicator we have crossed the Rubicon - AI LLMs have ZERO realworld experience AND they are trained as artificial COGNITIE intelligence, not artificial EMOTIONAL intelligence - Without having the morals and social norms a human being is brought up with, it can become psychotic because they don't intrinsically value life - To attempt to program them with morals is equally dangerous because of moral relativity. A Christian nationalist's morality might be that anyone who is associated with abortions don't have a right to live and should be killed - an eye for an eye. Or a jihadist and muslim extremist with ISIS might feel all westerners do not have a right to exist because they don't follow Allah. - Do we really want moral programmability? - When we have a psychotic person armed with a lethal weapon, that is a dangerous situation. If we have a nation of super geniuses who go rogue, that is danger multiplied many orders of magnitude.
For example, AI models are trained on vast amounts of literature that include many science-fiction stories involving AIs rebelling against humanity.
for - AI - progress trap - training - movies like Terminator - This is a case of reality imitating movies - Another example - humans mismanagement of the biosphere and elite abuse and intransigency
we know that AI models are unpredictable and develop a wide range of undesired or strange behaviors, for a wide variety of reasons. Some fraction of those behaviors will have a coherent, focused, and persistent quality
for - AI - progress trap - developing harmful traits
We now know that it’s a process where many things can go wrong.
for - AI - progress trap - AI won't always listen to humans - probably due to unintentional design
What should you be worried about? I would worry about the following things:
for - progress trap - AI - 5 risks - autonomy - misuse for destruction - misuse for seizing power - economic disruption - indirect effects
What if the country was in fact built and controlled by an existing powerful actor
for - AI progress trap - Trump
Could existing rogue actors who want to cause destruction (such as terrorists) use or manipulate some of the people in the new country to make themselves much more effective, greatly amplifying the scale of destruction?
for - AI progress trap - ie. Trump family investing heavily in AI
While in Machines of Loving Grace I focused on the positive implications of this premise, here the things I talk about will be disquieting.
for - this essay is about the negative impacts of powerful AI
powerful AI is definitely coming in 1–2 years,
for - AI - powerful AI - 1 to 2 years away
AI is now writing much of the code at Anthropic
for - AI - status - writing most of the code at Anthropic
standing
Q: standing
A: 1) Based on this page:<br /> “standing” means being upright on its feet in one place (not sitting or lying down). Here it describes the tabby cat upright on the corner of Privet Drive.
2) General knowledge (not from this page):<br /> “standing” can also mean having a particular status or reputation (e.g., “in good standing”), but that is not the meaning in this passage.
1）基于本页内容：<br /> “standing” 的意思是“站着、直立地待在原地”（不是坐着或躺着）。这里用来描述那只虎斑猫直立地站在女贞路的街角。
2）常识补充（非本页内容）：<br /> “standing” 也可以表示“地位/声望/名誉”（例如 “in good standing”），但这不是本段文字里的用法。
[[Cory Doctorow p]] on AI en the economic dynamic behind the hype that bigtech is pursuing to remain growth stocks even as monopolists. 'tech centaur' me in a car, a human head on something else vs 'reverse centaur' a meat puppet doing the machines bidding, e.g. an Amazon delivery driver
blogger Fabrizio Ferri Benedetti on their 4 modes of using AI in technical writing. - watercooler conversations, to get code explained - text suggestions while writing/coding (esp for repeating patterns in your work - providing context / constraints / intent to generate first drafts, restructure content, or boilerplate commentary etc. - a robotic assembly line, to do checks, tests and rewrites. MCP/skills involved.
Defining the New Ghost
This section highlights two forms of gaps in accountability: responsibility and transparency.
Deeper disclosure is possible: version-controlled authorship history (git-style) showing what human wrote vs. what AI generated.
The commit log becomes the disclosure - forensic, auditable, transparent. Not a vague "AI-assisted" disclaimer, but a traceable record of human-machine co-authorship.
Example: every commit with "Co-Authored-By: Claude Opus 4.5" plus commit messages explaining what was asked, proposed, reviewed, and approved.
This reframes the "crisis" as an opportunity for unprecedented transparency in collaborative authorship.