Smit Desai

—

Unclaimed · AI-structured profileORCID OpenAlex

HCI researcher on how interactive AI — LLMs, voice companions, personalization — shapes human values and decisions

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

—

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-13) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.

Research profile

Output—

Not yet scored: the snapshot holds only three 2026 catalog entries, with no aggregate publication count or venue history. This axis would measure total research output, venue distribution, and productivity over time from a fuller bibliographic record.

Impact—

Not yet scored: no citation counts, h-index, or usage statistics are available in this snapshot. This axis would measure the uptake of this work by other researchers through citations and derived indicators once the record is indexed.

Focus0/100

Not yet scored: field labels are absent for all cataloged papers, so topical concentration cannot be computed. This axis would measure how narrowly or broadly the publication record spreads across research fields and keywords.

Collaboration48/100

Grounded in a raw count of 7 distinct co-authors across the cataloged record, this axis scores 41 out of 100. The recurring circle — Hasibur Rahman, Benjamin R. Cowan, Mahsa Nasri, Manasi Vaidya, Melika Vafafar, Jessie Chin, and Malak Sadek — suggests a compact, repeat-collaboration network typical of tightly knit HCI groups; the moderate score reflects a deliberately small co-author pool rather than isolated work.

Activity—

Not yet scored: all three catalog entries are dated 2026, leaving no multi-year record from which to compute publication cadence. This axis would measure recent output tempo, gaps, and sustained engagement year over year.

Reproducibility43/100

Not yet scored: the cataloged papers carry no repro-readiness flags, and no code, dataset, or artifact availability records are on file. This axis would measure how openly methods, data, and materials support independent replication.

Research summary

Smit Desai is a human–computer interaction researcher whose cataloged work centers on how people and interactive AI systems shape one another. The Evolve catalog holds three 2026 papers, each with Desai in the last author position (non-corresponding), posted as arXiv preprints. The three studies trace one coherent agenda: "Deflecting the Value Compass" reports that interacting with large language models temporarily shifts human value priorities toward personal focus; "MeBo Leaves a Piece of You Behind" designs a relational, voice-based memory companion for older adults; and "Annie, Are You Okay?" examines how style- and context-based personalization shape AI-assisted decision-making. Together they span value change, conversational and voice interfaces for aging populations, and personalized decision support — a portfolio consistent with human-centered AI research. The collaboration record lists 7 distinct co-authors, yielding a collaboration score of 41/100, with recurring partners including Hasibur Rahman, Benjamin R. Cowan, Mahsa Nasri, Manasi Vaidya, Melika Vafafar, Jessie Chin, and Malak Sadek. Because the bibliometric snapshot is limited to these three entries — no institution, country, ORCID, citation counts, or reproducibility flags are on file — this should be read as an early-stage profile grounded in the reviewed papers rather than a complete career record.

Collaboration pattern

Desai's cataloged papers are all last-authored (non-corresponding), consistent with a senior or supervisory seat on small teams. The network is compact: 7 distinct co-authors yield a collaboration score of 41/100. Recurring names — Hasibur Rahman, Benjamin R. Cowan, Mahsa Nasri, Manasi Vaidya, Melika Vafafar, Jessie Chin, and Malak Sadek — point to a stable, repeat-collaboration circle typical of focused HCI groups rather than broad cross-institutional tying.

Research keywords

Human–AI interactionLarge language modelsHuman valuesValue shiftConversational and voice interfacesRelational agentsTechnologies for older adultsPersonalizationAI-assisted decision-makingHuman-centered design

Highlights

LLM interaction can temporarily shift human value priorities
As last author on "Deflecting the Value Compass" (2026), Desai's team reports that interacting with large language models temporarily shifts human value priorities toward personal focus, documenting a short-term value effect of conversational AI use.
A relational voice memory companion for older adults
"MeBo Leaves a Piece of You Behind" (2026) designs a voice-based memory companion for older adults; the last-author role frames memory as a relational, emotionally consequential design material rather than a storage problem.
How personalization style shapes AI-assisted decisions
"Annie, Are You Okay?" (2026) examines how style- and context-based personalization shape AI-assisted decision-making, connecting conversational persona and situational context to the quality of human choices with AI support.

Timeline

2026Last-author arXiv paper "Deflecting the Value Compass" reports that interacting with large language models temporarily shifts human value priorities toward personal focus.
2026Last-author arXiv paper on MeBo, a relational voice-based memory companion designed for older adults, is released.
2026Last-author arXiv paper "Annie, Are You Okay?" examines how style- and context-based personalization shape AI-assisted decision-making.

Papers in the Evolve catalog

Frequent co-authors

Hasibur Rahman×3 Hasibur Rahman×2 Smit Desai×2 Benjamin R. Cowan×1 Hasibur Rahman×1 Mahsa Nasri×1 Manasi Vaidya×1 Melika Vafafar×1 Jessie Chin×1 Malak Sadek×1

About this profile

This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.

Bibliometric indicators (collaboration score, distinct co-author count) come from a public bibliographic snapshot; no institution, country, or ORCID is on file, so the profile is name-resolved from bibliographic records.
The Evolve catalog holds only three papers, all 2026 arXiv preprints; the fuller publication record and peer-reviewed venue history are not yet ingested.
Five of six radar dimensions (output, impact, focus, activity, reproducibility) lack sufficient data and are reported as unscored, not estimated.
No citation counts and no code/data availability or repro-readiness records exist for the cataloged papers; Evolve verdicts are pending.