A useful conversation starts with a precise question. This selected update pairs recent announcements with a study about professional learning. It is not a complete news digest or a Wisomi product test.

ChatGPT adds collaborative slides, with rollout limits

Publication/event: 2026-10-16 · Checked: 2026-10-17

OpenAI announced collaborative slide creation in ChatGPT on 16 October. Eligible Pro, Business and Enterprise users with Space access are receiving the rollout. Managers could evaluate a routine meeting deck using approved, non-sensitive material, checking its story, figures and exported layout. Include correction time when assessing usefulness. Wisomi verified the announcement but has not tested this feature.

Limits: Rollout and export quality untested by Wisomi; individual account availability unconfirmed

Source: OpenAI announcement (opens in a new tab).

Gemini 4 Argon moves beyond its initial restricted rollout

Publication/event: 2026-10-13 · Checked: 2026-10-17 · Initial announcement: 2026-09-30

Google expanded Gemini 4 Argon access on 13 October, after its restricted September introduction. Its announcement covers APIs and eligible Google AI Pro and Ultra subscribers in the Gemini app and AI Mode. Before relying on it, professionals could evaluate a non-sensitive task with a known answer, checking errors, review effort and cost. Wisomi verified the announcement, not the model's performance.

Limits: Wider availability does not mean free access; provider performance claims untested by Wisomi

Source: Google announcement (opens in a new tab).

Better AI-assisted work does not establish unaided skill

First issue: September 2026 · Manuscript: 7 October 2026 · Checked: 17 October 2026

The study compared AI-assisted drafting with a later assessment where AI use was prohibited. The AI-access group drafted better; its advantage on the later assessment was concentrated among lawyers with at least seven years' experience. No average advantage was detected for juniors. This Google-funded working paper covers one specialty: 91 of 133 participants completed the full protocol, with no baseline skills test. Wisomi proposes checking both assisted work and a later unaided attempt; that exercise was not tested here.

Limits: Working paper; Google-funded; 11 U.S. firms; no baseline skills test; potential non-adherence weakens full-sample significance

Source: Does AI Assistance Enhance or Erode Expertise? Evidence from a Three-Month Field Experiment in Patent Drafting (opens in a new tab).

This selected update contains original Wisomi summaries and an untested learning exercise. Product performance was not tested by Wisomi. Community posting and Workspace access are not open.