iiinova — where your digital self keeps evolving

Reviewer 3 reading my paper

The faces of Chlöe Swarbrick in parliament

1332

annals of book research, vol. 12, comparative psych edition in land-dwelling vertebrates, the animal holds still and the sound shakes the ear. in water-dwelling animals, the ear holds still while the sounds shake the fish

The Goldfish Ear Codes the Axis of Acoustic
Particle Motion in Three Dimensions
Abstract. Auditory and vestibular nerve fibers of the goldfish are strongly
directionally sensitive to whole-body acceleration at audio frequencies. The threedimensional pattern of sensitivity shows that input from a receptor ensemble (hair
cells) is essentially equivalent to that expectedfrom a single hair cell having a given
three-dimensional orientation of best sensitivity. Fibers from the sacculus, lagena,
and utriculus differ with respect to distributions of directional orientation, but are
similar in best threshold (less than 1 nanometer, root mean square, at '140 hertz). In
combination with other mechanisms for detection ofsound pressure, this directionality is a likely basis for directional hearing in fishes, and it could allow the determination of underwater acoustic intensity.

020

48. SHORT TALKS (Anne Carson). This is one of Carson's early books of poetry, or maybe better to call them vignettes, or short talks since that's what she called them For me, reading this was a bit like walking through a well-curated gallery without knowing much about the painters

010

lab managers: choose your participant incentives such that the participants are still using them thirty years later

À canvas bag that says COMET correction of myopia evaluation trial

3220

no shade on Carol but would it kill academic publishers to hire a cover designer now and then

The cover of COGNITIVE FOUNDATIONS OF MUSICAL PITCH, which is a blue rectangle on top of a red background

8563

in 2020 my lab quit Microsoft Word and began writing in R Markdown, an open-source document-publishing system that facilitates a clear, permanent link between statements in text (scientific prose) and the code producing analyses and visualizations supporting those statements (scientific analyses)

a preprint generated with R Markdown with the following text

Ecological momentary assessment reveals causal effects
of music enrichment on infant mood
Eun Cho1,∧,∗
, Lidya Yurdum1,2,∧,∗
, Ekanem Ebinne1
, Courtney B. Hilton3
, Estelle Lai3
, Mila Bertolo1,4,5
,
Pip Brown3
, Brooke Milosh6
, Haran Sened7
, Diana I. Tamir7
, and Samuel A. Mehr1,3,∗
1Child Study Center, Yale University, New Haven, CT 06520, USA.
2Department of Psychology, University of Amsterdam, Amsterdam 1018WT, Netherlands.
3School of Psychology, University of Auckland, Auckland 1010, New Zealand.
4Centre for Research in Brain, Language and Music, McGill University, Montréal, QC H3G 2A8, Canada.
5
Integrated Program in Neuroscience, McGill University, Montréal, QC H3A 1A1, Canada.
6Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, NY 11549, USA.
7Department of Psychology, Princeton University, Princeton, NJ 08544, USA.
∧These authors contributed equally.
*Corresponding authors. E-mails: eun.cho@yale.edu, lidya.yurdum@yale.edu, sam@auckland.ac.nz
Music appears universally in human infancy with self-evident effects: as many parents know intuitively,
infants love to be sung to. The long-term effects of parental singing remain unclear, however. In an offsetdesign exploratory 10-week randomized trial conducted in 2023 (N = 110 families of infants, Mage =
3.67 months, 53% female, 73% White), the study manipulated the frequency of infant-directed singing
via a music enrichment intervention. Results, measured by smartphone-based ecological momentary
assessment (EMA), show that infant-directed singing causes general post-intervention improvements
to infant mood, but not to caregiver mood. The findings show the feasibility of longitudinal EMA
(retention: 92%; EMA response rate: 74%) of infants and the potential of longer-term and higherintensity music enrichment interventions to improve health in infancy.
Keywords: music, infancy, parenting, infant-directed song, ecological moment…

the journal article version of the previous image

518247

a couple ways I serve hummus when not eating it straight out of the dish standing in front of the fridge, possibly with a spoon, are: (L) toasted pine nuts, good olive oil, paprika or (R) warm with grilled beef tenderloin chunks, toasted pine nuts, parsley and always with fresh pita

a bowl of hummus with pine nuts and evoo and paprika, next to pitas

hummus in a big bowl with pine nuts beef and parsley

2120

here is my hummus recipe since it is perfect (certainly better than Yotam's which it's adapted from)

Hummus

10 Oct 2026 Sam
Adapted from Ottolenghi to be smaller and less of a pain to make. This recipe starts with enough chickpeas for 2 batches of hummus and only uses half of them (freeze the other half so that you can go right to step 3 next time).

250g dry chickpeas
1 tsp baking soda
2-3 litres water
150g light tahini paste
3 Tbsp lemon juice
2-3 cloves garlic, smashed
1 tsp table salt
50g ice

1. Prep the chickpeas
Soak 250g dry chickpeas in cold water overnight or up to 24 hours. Discard any floaters.

2. Cook the chickpeas
Drain the chickpeas and put them in a large pot with a teaspoon of baking soda on high heat. Stir for three minutes, then add 2-3 litres of water and bring to a boil. Simmer for ~30 min, skimming with a spider regularly to first remove scum and then skins as they blanch off. The chickpeas are done when they fall apart between two fingers. Drain the chickpeas and you should be left with 575-600g of wet chickpeas.

Put half in a small cuisinart (if using) and freeze the other half for a future batch of hummus. You can also do everything in a pot with a stick blender, or a double or triple batch in a Vitamix if you want it perfectly smooth.

3. Make the hummus
Process 275-300g of chickpeas, still warm, until they form a thick paste (if using from frozen, thaw and warm in microwave before starting). Add everything else and process for 4-5 minutes until it's ultra-smooth. If it's a bit thick you can lighten with lemon juice, water, olive oil, or all three. Transfer to a dish, cover surface with cling film so that it doesn't form an unpleasant skin as it cools, and refrigerate. Serve at room temp.

7588

Our sparkie does find this funny in an Auckland kitchen though lol

040

I love my mini one, it's from the days when they built them with a beefy motor despite the small footprint. The newer ones are garbage in comparison, imo (with the exception of commercial robot coupe sticks like these)

220

this is the only brand I've found in Auckland that's available cheaply in decent quantities, although I bet there are more options if I go to the middle eastern market on the other side of the city (this is from the restaurant supply place closer to us)

110

I fully intend for my Little Pro Plus to outlive me; our larger Cuisinart is older than i am (born Dec 1978 if we are to believe the stamp on the bottom) and probably will also outlive me

the stamp on the bottom of our Cuisinart, made in Le Perreux, France in 1978

490

I have made hummus in various appliances and while the smoothest results come from the vitamix, it requires a quite large batch size, more than I want sitting in the fridge. so I use my 1990s-era Cuisinart Little Pro Plus which looks like this and it is perfect

cuisinart little pro plus with its box and cooking book

2120

sometimes on Saturday morning i make hummus

2360

i stg every time i put something in the microwave I am reminded of the biteNUKER

180

Dinner was burgers and chopped salad. We eat with our eyes too: colours are important.

À mess of a double smash burger with lots of veg

1160

fave new demonstration of survey items being really not that bad at tracking carefully-measured cognitive things: in 900,000 participants, a 5-item "how good are you at music" self-report (the colours) very directly tracks a 20min adaptive test of music perception ability (the x-axis)

a series of histograms of self-assessed music skill lined up very cleanly with music perception ability

88418

adoption of AI tools leads to substantially more code, commits, and pull requests — but (thanks to v rich dataset) when they directly assess how this translates to productivity, the story is v different the effect of AI adoption in resolution of Jira tickets & epics is not distinguishable from zero

lots more code, commits, and pull requests when adopting AI agents

no concurrent increase in jira issues or epics resolved

3154

the dataset is pretty kickass: 300 million 'work events' (code commits, pull requests, jira tickets, etc) from 700,000 employees at 700 companies, which each adopted vibe-coding tools at different times, enabling a difference-in-differences study

We use a novel proprietary dataset from Jellyfish (JF), a software firm whose main product is an
analytics platform for companies to understand the activity of their engineering teams. Our data
provides information on approximately 300 million work events of 700,000 workers across 700 firms,
from January 2021 to March 2026. JF integrates data from five sources: (1) HR information; (2) code
version control systems, such as GitHub, which provide information on engineers’ coding activity;
(3) issue management systems, such as Jira, which indicate engineers’ assignment to and progress
on tasks and projects; (4) scheduling tools, such as Google Calendar, which provide information on
2
meeting activity and time allocation; and (5) generative AI usage, for GitHub Copilot, Cursor, and
Claude Code. The scope of our data allows us to observe the full software production pipeline, from
intermediate measures of coding activity to final software output, as well as employment, and to tie
changes in these outcomes to AI adoption.

1111

this working paper from Harvard PhD students Fiona Chen and James Stratton is making the rounds, perhaps because it confirms many people's suspicions about the tradeoffs in vibe coding: LLMs mean lots more code written, but also lots more problems created, and relatively little productivity benefit

Artificial Intelligence in the Firm:
Bottlenecks in Software Production
Fiona Chen
Harvard University
Job Market Paper
James Stratton
Harvard University
Current version: August 4, 2026
First version: January 7, 2026
Click here for current version
Abstract
We study the impacts of AI coding assistants and agents on software engineering work, using
a novel proprietary dataset from an engineering analytics platform, covering 300 million work
events — including GitHub coding activity, Jira issues, and Google Calendar events — across
718 firms. We use a staggered difference-in-differences design, exploiting variation in firmlevel adoption timing of AI coding assistants and agents. Both technologies increase coding
productivity. However, productivity gains do not fully pass through to changes in software output
or employment. For AI agents, this incomplete pass-through reflects a bottleneck from code
review: review times increase, a larger share of code updates require revisions, and reviews
involve more comments. We develop a model of software production to structure these results, in
which AI affects both productivity and quality of intermediate outputs, and in turn can generate
a review bottleneck and limit pass-through.

97021