The OS powering Assistantly
All issues

Intelligence Brief

Week of
October 12, 2026

Your weekly edge, the research, frameworks, and signal that separate operators from spectators. Read and summarized so you do not have to.

Published Monday, October 12 at 6:00 AM PT. 12 sources, 8 minute read.

01

The Human Intelligence Edge

What makes humans irreplaceable in the AI era.

IBM Institute for Business ValueSeptember 21, 2026

New IBM CHRO Study: AI Puts Critical Thinking at the Center of Workforce Priorities

IBM Institute for Business Value with Oxford Economics

Why it matters. Across 1,500 CHROs and 8,800 employees, 71% of the CHROs named the ability to supervise, validate, and override AI as a top workforce capability, while 60% of the employees said they worried AI was eroding their skills. Those are the same observation taken from two heights. The capability executives now want most is the one their own people believe the daily work is wearing away. Note what the capability actually is: not fluency with a model, but outranking it on a specific call and being able to say why. That is testable in an hour. Hand someone a confident, wrong output inside their own domain and watch whether they take it.

Read the source
arXiv preprintSeptember 28, 2026

Cognitive Skills in the Age of AI: Computing Students and Experts Perceptions

Neha Rani, Vu Minh Anh Le, Austin M. Spangler, and Erta Cenko, IEEE ICALT 2026

Why it matters. Students and experts in this mixed-methods study expect many cognitive skills to lose importance in an AI-rich workplace, with critical thinking the exception they keep. The paper does not state a sample size, so read the shape of the answer rather than its weight. What it records is a decision already made by the people entering these jobs about which of their own abilities are worth maintaining, and that belief governs what they practice whether or not it turns out to be right. A cohort optimizing for the skills it expects to survive will arrive fluent in tools and thin on the reasoning that checks them, unless something in the hiring process asks for the second thing out loud.

Read the source
NatureSeptember 23, 2026

How to stay smart in the age of AI: the science of critical thinking

Helen Pearson

Why it matters. Pearson's headline numbers are forecasts from the supply side: 90% of more than 1,000 surveyed US faculty expect AI to reduce students' critical thinking, and about one in five AI-using students in a separate survey of more than 8,000 say working a problem without it has gotten harder. Set that beside the 71% of CHROs in the IBM study and both halves of a shortage are now on record, one in the hiring plan and one in the classroom. Neither side is causal evidence, and both are firm statements of expectation from the two groups that set supply and demand. The working conclusion for a hiring team is that this capability will not arrive attached to a degree, and it does not stay fresh on its own once someone is inside your company either.

Read the source
02

The Broken System

Why resumes, interviews, and hiring infrastructure are failing.

International Journal of Selection and AssessmentJuly 9, 2024

Evaluating interview criterion-related validity for distinct constructs: A meta-analysis

Timothy G. Wingate, Joshua S. Bourdage, and Piers Steel

Why it matters. Across 37 studies and 30,646 participants, interview assessments related moderately to both task and contextual performance, and the ratings did not cleanly distinguish the two. The interview predicts something real. It does not rate the separate things your scorecard says it is rating, which is what happens when one overall impression gets written into five boxes. What survives the meta-analysis is structure: job-relevant questions and explicit scoring criteria. Anything a free conversation adds past that is mostly rating the conversation, and thirty thousand participants is a sturdier reason to rebuild your interview than any argument about AI.

Read the source
SHRMSeptember 15, 2026

Fraudulent Candidates at Scale: Are AI-Generated Applications Breaking Recruitment?

SHRM Advisor, citing Gartner survey data

Why it matters. Both Gartner figures here, 39% of 3,290 candidates using AI in applications and 6% of a separate 3,000 admitting to interview fraud, were reported by the people doing it. Self-reported rates for behavior you are not supposed to admit are floors, not estimates. The consequence is not a tighter policy, because a rule nobody can verify only binds the candidates honest enough to follow it. It is that assistance is now the default condition of an application, which makes the question worth asking in the room a different one: what did you use it for, and where did you disagree with what it gave you? That beats most of the questions it would replace.

Read the source
arXiv preprintSeptember 17, 2026

When Hiring Becomes Agent-Mediated: Evaluating Access and Recurrence in Two-Agent Resume Screening

Jian Gao and Hang Jiang, accepted to the REALM Workshop at EMNLP 2026

Why it matters. The part to sit with is not that two-agent screening advanced a different set of candidates than a single scoring call across 600 constructed pairs. It is that the result moved with the model and with the run. A screen that answers differently on a rerun is not a measurement, and no hiring team would accept that variance from a human panel or defend it to the person it rejected. These are simulated pairs rather than real hiring outcomes, so treat the instability as a property of the method and not as a published rejection rate. Then rerun your own first pass over last month's pipeline and see what moves.

Read the source
03

The SupaHuman Standard

The behavioral science of who actually thrives.

DeloitteJanuary 14, 2026

Human capabilities are at the heart of high-performing teams

Deloitte, survey of 1,394 employees

Why it matters. This is the oldest source in the issue, which is the reason to read it again now. In January it put curiosity, resilience, informed agility, connected teaming, and emotional and social intelligence under high-performing teams, and found those teams using AI at 78% against 54% for everyone else. Nine months later MGI finds employers writing the same words into job postings. The association proves no direction and the sequence is still worth noticing: these went from a consultancy's list of capabilities to a stated hiring requirement inside a year, which is faster than any credential system can answer.

Read the source
Boston Consulting GroupSeptember 29, 2026

The Three Leadership Roles CEOs Need for AI in 2026

Brittany Heflin, Gabriella Kellerman, Robert Werner, David Martin, Julia Dhar, and Vladimir Lukic

Why it matters. Across 563 global executives, those at AI-mature organizations were 75% more likely to experiment, seek feedback, and adapt, and organizations combining leadership readiness with mature AI practice reported an 80% lift in competitive performance. Both figures are self-reported and correlational, so the 80% is not the number to carry out. The list of three behaviors is. Every one of them is observable and none of them is a trait: you can watch a person run an experiment, ask for a critique they did not want, and move their position inside a single work trial. That is more than a leadership competency framework has ever shown you about a candidate.

Read the source
McKinsey Global InstituteSeptember 29, 2026

Workforce in motion: Skills and pathways to future jobs in the United States

María Jesús Ramírez, Kweilin Ellingrud, Tanguy Catlin, Diego Castresana, and Anna Kortis

Why it matters. Adaptability demand across US job postings rose about fivefold from 2022 to 2026, with resilience, curiosity, and willingness to learn roughly tripling. That is a price signal, not a performance finding. It records what employers will now commit to in writing and says nothing about who actually holds the trait, which sets up the next problem: once every posting asks for adaptability, the word filters nobody, and candidates write it back into every resume within a quarter. The only version of the requirement that keeps working is an observable one, a specific account of someone's brief changing under them and what they did in the week after.

Read the source
04

The Future of Work, Without the Clichés

Where AI and human collaboration is actually heading.

McKinseyOctober 5, 2026

AI is changing work. Now it has to change the organization

Alexis Krivkovich, Brooke Weddle, Holly Price, and Vik Sohoni

Why it matters. 80% of respondents said AI improved their individual productivity while only 6% of their organizations qualified as AI high performers, and nearly three-quarters of those high performers had substantially redesigned workflows against about a quarter of everyone else. The distance between 80 and 6 is the finding. Individual gains stay individual until the work around them is rebuilt, and rebuilding it is an organizational act rather than a purchase. For hiring that changes what a strong candidate looks like in a role whose shape is about to move: someone who can tell you how their own work should be reorganized is worth more than someone who is merely fast inside the current version of it.

Read the source
Boston Consulting GroupSeptember 24, 2026

Five Ways That AI Front-Runners Change How Work Gets Done

Sagar Goel, Julie Bedard, Matthew Kropp, Ashley Sim, and Charikleia Kaffe

Why it matters. Interviews across 50 companies keep returning the same division of labor: AI produces, people frame the problem, direct the work, and judge the result, and a team owns the outcome. Look at which half of the job that leaves to people. Framing and evaluating were never entry-level work, and the task layer being automated is the layer that used to buy someone their first five years. BCG's association between work and talent transformation and more than 11 percentage points of higher annual industry-adjusted shareholder return rests on qualitative interviews, so leave it where it sits. The design question underneath it is real and it lands on hiring: if the junior work goes, you are recruiting for judgment at the entry point and you need a way to see it that does not assume a track record.

Read the source
BCG Applied AI Index 2026September 30, 2026

The Formula for Agentic AI Value

Jessica Apotheker et al., survey of more than 1,300 senior leaders

Why it matters. Among more than 1,300 senior leaders across 20-plus sectors, 89% expect AI to generate new work and 11% expect it mainly to replace existing jobs, while strategic workforce planning runs at 55% among future-built companies against 17% for laggards. Read the first number as what it is, an expectation, and leaders are not neutral narrators of their own headcount plans. The second is a behavior, which makes it the better signal. Working out which human roles the new workflows will need is the scarce act here, and the companies doing it are not forecasting the transition. They are staffing it, which is a different and much earlier job.

Read the source
Previous issueWeek of October 5, 2026 All issuesBrowse the archive