cd /news/generative-ai/practitioners-use-ai-at-execution-la… · home topics generative-ai article
[ARTICLE · art-16527] src=letsdatascience.com pub= topic=generative-ai verified=true sentiment=· neutral

Practitioners Use AI at Execution Layer, Judgment Matters

A Drexel University study of 205 real-world ChatGPT use cases found that 47% of interactions involved writing tasks, with identifying accounting for 10%, according to reports from Search Engine Journal and Duane Forrester's Substack. The research, conducted by Tim Gorichanaz, identified six usage modes—writing, deciding, identifying, ideating, talking, and critiquing—with data drawn from Reddit and skewed toward Anglophone users. The findings indicate that most practitioners apply generative AI at an execution layer for text creation rather than higher-value judgment tasks, a pattern that affects career differentiation and strategic leverage for teams.

read3 min publishedMay 28, 2026

Search Engine Journal and Duane Forrester's Substack summarise a Drexel University study by Tim Gorichanaz that analysed 205 real-world ChatGPT use cases and identified six usage modes: Writing, Deciding, Identifying, Ideating, Talking, and Critiquing. The study found Writing accounted for 47% of observed cases and Identifying about 10%, and Search Engine Journal reports the dataset came from Reddit and skews Anglophone. Search Engine Journal also cites a figure that 63% of organisations using generative AI apply it primarily to create text. Editorial analysis: this concentration on drafting and factual synthesis means many practitioners are using AI at an execution layer rather than the higher-value judgment layer, with implications for career differentiation and where teams extract strategic leverage.

What happened

Search Engine Journal and Duane Forrester's Substack summarise a Drexel University paper by Tim Gorichanaz that analysed 205 real-world ChatGPT use cases and produced a six-mode taxonomy of how people actually use conversational generative AI. The six modes the paper identifies are Writing, Deciding, Identifying, Ideating, Talking, and Critiquing. Per the Drexel dataset as reported by Search Engine Journal, Writing comprised 47% of observed uses and Identifying comprised 10%. Search Engine Journal additionally reports the study's cases were drawn from Reddit and skew Anglophone, and cites a separate enterprise figure that 63% of organisations using generative AI apply it primarily to create text.

Editorial analysis - technical context

The taxonomy separates work that automates execution (drafting, summarising, translating) from work that requires human judgment (evaluating tradeoffs, forming strategy, nuanced critique). Industry-pattern observations: practitioners and organisations often optimise for near-term productivity gains by applying models to repeatable text tasks, which increases throughput but concentrates value in routine outputs rather than decision-making nodes.

Context and significance

For practitioners, concentration in the Writing and Identifying modes reduces the marginal upside of tooling improvements aimed solely at execution. Industry-pattern observations: when a workforce primarily applies AI to execution-layer tasks, strategic differentiation shifts to roles that integrate model outputs into higher-order judgment, such as framing problems, adjudicating model error modes, and synthesising ambiguous signals across domains.

What to watch

Metrics and signals an observer should follow include broader usage surveys that break down modes beyond content creation, hiring and role postings that emphasise decision-support or interpretability skills, and tool features that surface provenance, counterfactual reasoning, or critique workflows rather than only faster drafting. Industry-pattern observations: rising demand for tooling and processes that make model outputs auditable and deliberative would indicate movement from execution-layer adoption toward judgment-layer workflows.

Scoring Rationale #

The Drexel study surfaces a widely observable pattern in practitioner AI use that affects how teams capture value. It is notable for user-research insights relevant to practitioners but not a frontier technical breakthrough, and it is recent (within days), yielding a modest downward freshness adjustment.

Practice interview problems based on real data

1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.

Try 250 free problems

── more in #generative-ai 4 stories · sorted by recency
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/practitioners-use-ai…] indexed:0 read:3min 2026-05-28 ·