Matt Garman, the chief executive of Amazon Web Services (AWS), told CNBC that in the coming AI age the most valuable talents won’t be narrow technical chops like coding but critical thinking, creativity, adaptability and strong communication skills—a message he says he gives his own children and one that reframes how higher education and employers should prepare people for work.
Matt Garman took the reins of AWS with deep experience inside Amazon and has overseen a strategic pivot that places artificial intelligence at the center of the cloud business. AWS today sells everything from foundation models to managed AI platforms such as Amazon Bedrock while investing heavily in custom silicon and data-center capacity to power large-scale model training and inference. Those investments are part of a broader company calculus that treats AI as an operational and economic lever. At the corporate level Amazon’s leadership has been explicit that generative AI and agentic tools will change the shape and size of white‑collar work. CEO Andy Jassy has warned employees that as the company rolls out AI agents, some roles will require fewer people and others will change substantially—an acknowledgement that automation will reallocate tasks across teams and job families. That wider corporate posture provides the context for Garman’s advice: AWS is building the capability to automate more of the routine and administrative work, and company leaders are increasingly focused on how humans add value where machines do not.
That said, the prescription is not a binary choice. The highest‑value profiles will fuse durable human skills with domain competence and technical literacy. Stakeholders—universities, employers, policymakers and technologists—should treat Garman’s advice as a strategic orientation: invest in human critical faculties while building tools, curricula and policies that help large populations acquire them. Companies must align their reskilling investments and workforce transition plans with the rhetoric, or risk eroding the very trust they need to navigate a decade of rapid change.
Garman’s short message to his children—“develop your critical thinking skills in college, no matter what subject you study”—is a practical, portable guideline for a volatile labor market. The work ahead is institutional: to make that advice actionable and equitable at scale, stakeholders must redesign learning pathways, measurement tools, and corporate transition policies so that people, not just machines, capture the gains of the AI era.
Source: Storyboard18 AWS Chief reveals list of skills other than coding that will define the AI age
Background
Matt Garman took the reins of AWS with deep experience inside Amazon and has overseen a strategic pivot that places artificial intelligence at the center of the cloud business. AWS today sells everything from foundation models to managed AI platforms such as Amazon Bedrock while investing heavily in custom silicon and data-center capacity to power large-scale model training and inference. Those investments are part of a broader company calculus that treats AI as an operational and economic lever. At the corporate level Amazon’s leadership has been explicit that generative AI and agentic tools will change the shape and size of white‑collar work. CEO Andy Jassy has warned employees that as the company rolls out AI agents, some roles will require fewer people and others will change substantially—an acknowledgement that automation will reallocate tasks across teams and job families. That wider corporate posture provides the context for Garman’s advice: AWS is building the capability to automate more of the routine and administrative work, and company leaders are increasingly focused on how humans add value where machines do not. What Garman actually said — and why it matters
Garman’s concise counsel to students and early-career workers is threefold: develop critical thinking, cultivate adaptability, and hone communication. He framed higher education less as vocational training and more as a crucible for reasoning and judgment—skills he described as the likely differentiator in an environment where AI can perform many task-level functions. In the interview he told CNBC that “develop your critical thinking skills in college, no matter what subject you study,” and called critical thinking “the number one, biggest key to success in the age of AI.” Why this matters: senior AWS leaders are not just philosophizing. They are steering a cloud provider that sells AI tooling to enterprises and is simultaneously adapting its own workforce and product roadmaps to the implications of those tools. When executives at scale underscore durable human capabilities, it signals both market demand and a practical response to technological limits. That signal reverberates across education, talent marketplaces, recruiting practices, and the investments companies make in reskilling programs.The case for soft skills: evidence and nuance
AI is getting capable — but not omniscient
Generative models and agentic systems have advanced rapidly and can already automate many administrative, repetitive, and pattern‑matching tasks. That reality fuels corporate interest in using AI to improve efficiency. But state‑of‑the‑art models still struggle with reliably replicating higher‑order human faculties such as calibrated multi‑step reasoning, contextual judgment, and emotional nuance—areas where hallucinations, brittle reasoning and opaque failures persist. For roles that require moral judgment, complex operational tradeoffs, or customer empathy, current AI systems are imperfect partners at best. Multiple technical assessments show that hallucinations—confident but incorrect outputs—remain an unresolved challenge for large language models, even as accuracy improves on many benchmarks. Research papers and company technical notes stress that hallucination is rooted in the statistical structure of next‑token prediction and in incentive structures used during training; mitigation is possible but incomplete. That technical gap helps explain why employers will continue to prize human oversight, judgment and contextual awareness.Employers already list soft skills as core demand
Labor‑market analytics and industry reports reinforce the point. LinkedIn and other platforms consistently rank communication and adaptability among the most sought‑after competencies across industries; employers report that human skills like problem‑solving, relationship management and strategic thinking remain scarce and valuable. The data show that while technical skills fluctuate in importance as tools change, these human‑centered abilities are deeply transferrable and durable across roles. That empirical pattern aligns with Garman’s prescription.Soft skills complement rather than replace technical skills
It’s important to clarify what “soft‑skills first” advice does not mean: it’s not a call to abandon technical literacy. The most resilient career profiles in an AI‑infused economy are hybrid: people who combine domain expertise with high emotional intelligence, critical judgment, and adaptability. For example, site reliability engineers, product managers, and enterprise architects who can interpret model outputs, design guardrails, and communicate implications to non‑technical stakeholders are more valuable than practitioners who can only write code but cannot contextualize or defend its use. Employers will prize people who can coordinate human and machine workstreams.What higher education and training should do differently
Rebalance curricula toward reasoning and applied judgment
Garman’s recommendation that college should build reasoning and judgment suggests concrete curriculum shifts: emphasize case‑based learning, Socratic seminars, ethical reasoning, and multi‑disciplinary projects that force students to weigh tradeoffs and defend decisions. These modes cultivate the reflexes of critical thinkers: evidence evaluation, question framing, and intellectual humility.- Integrate cross‑disciplinary projects that pair technical builds with human‑centred evaluation.
- Use real‑world case studies that require students to defend tradeoffs under uncertainty.
- Teach source verification and information literacy so graduates can spot machine hallucinations.
Scale affordable, modular reskilling for workers
Employers should expand micro‑credentials, apprenticeships, and employer‑sponsored bootcamps that mix technical tool use with communication and judgment training. Companies including cloud providers are already funding training credits and learning pathways; a thoughtful program pairs hands‑on AI tool use with role‑play, feedback loops, and customer‑facing simulations to grow both skill sets at once.Strengths of Garman’s stance
- Practical realism: By centering skills that remain hard for machines, Garman avoids techno‑utopian determinism and focuses on what people can control: how they think and relate to others.
- Actionable guidance: Advising students to learn how to think — not just what to code — is portable, low‑cost advice that can be applied across disciplines.
- Alignment with labor demand: Data from talent marketplaces and corporate hiring teams corroborate the premium on communication and adaptability, giving the advice market validity.
Risks, blind spots and unintended consequences
Risk 1 — Oversimplification and messaging hazard
Saying “don’t code, learn critical thinking” risks being interpreted as a binary choice rather than a synthesis. For many roles technical depth remains essential. Presenting soft skills as a substitute for technical competence could mislead students into underinvesting in the domain knowledge employers still require. Educational messaging must emphasize both/and.Risk 2 — Uneven access to the “human edge”
Not all learners receive the same exposure to pedagogies that build judgment—elite institutions are more likely to offer seminar‑style courses, internships and mentoring. If employers reward those meta‑skills without broadening access, inequality could widen as students from advantaged backgrounds receive the soft‑skill training that others lack. Policy and philanthropic investments will be necessary to scale equitable access.Risk 3 — Corporate signaling vs. operational reality
Executives simultaneously champion soft skills and pursue aggressive automation and headcount reductions. That mixed signal—advise workers to develop judgment while changing role headcounts—can erode trust unless companies pair restructuring with transparent upskilling, fair transition plans, and meaningful internal mobility. Amazon’s own public statements about workforce shifts illustrate this tension: while leaders emphasize reskilling, they also acknowledge future reductions in some corporate roles.Risk 4 — Measurement, hiring and bias
Soft skills are harder to measure and standardize than coding tests. If hiring managers lean on subjective interviews or cultural fit proxies, assessment can amplify bias and reduce diversity. Organizations will need structured behavioral assessments, situational judgment tests, and transparent rubrics to evaluate critical thinking and communication without reproducing exclusionary patterns.Practical steps for IT professionals and Windows‑era technologists
Garman’s counsel is relevant to system administrators, DevOps engineers, Windows platform specialists, and IT managers who will work alongside AI tools. The pragmatic path forward is hybridization: deepen domain knowledge while deliberately practicing higher‑order human skills.- Learn to interpret AI outputs. Practice critiquing model responses, tracing failure modes, and documenting caveats.
- Develop clear stakeholder narratives: translate technical tradeoffs into business risks and customer impacts. Good communication multiplies technical value.
- Build adaptability routines: short sprints to learn new tools, playbooks for tool evaluation, and a personal learning ledger (courses, projects, reflections).
- Use role‑play and tabletop exercises to strengthen judgment in incident response, security tradeoffs, and product decisions.
- Cross‑train on adjacent disciplines: get comfortable with UX principles, project leadership, and basic statistics to make data‑informed choices.
How companies should operationalize the advice
Design roles that pair machines and humans
Job design should specify which parts of work are delegated to AI and which require human judgment. That clarity enables targeted hiring and internal mobility. Look for role definitions such as “AI‑augmented analyst” or “human‑in‑the‑loop operator,” which formalize the hybrid responsibilities.Invest in evaluative training and feedback loops
Training should be experiential and include structured feedback. For example, put employees through simulations where they must detect model hallucinations, defend a decision to a non‑technical stakeholder, and iterate based on customer feedback.Create transition safeguards
When workforce reductions occur, companies must offer generous transition support: internal redeployment, certified training tracks, and measured severance with job‑search assistance. Transparent timelines and concrete reskilling commitments reduce the social harm of change. The public record shows Amazon and other companies predicting workforce shifts while also pushing learning programs—execution matters.Policy and societal considerations
The economic shift described by Garman is not just a corporate HR problem; it raises public policy questions. Governments and educators must:- Reassess financing for lifelong learning programs and micro‑credentials.
- Encourage assessment standards for soft skills that protect against bias.
- Fund regional reskilling initiatives tied to employer demand.
- Strengthen unemployment‑to‑retraining pathways so workers displaced by automation can transition into newly valuable roles.
Final analysis: a sensible reframing with caveats
Matt Garman’s message is a useful corrective to a narrowly technical career narrative: as AI automates certain tasks, the human advantages will be the capacities that are hardest for machines to replicate—judgment, adaptability, and human connection. The claim is supported by labor‑market signals and by the technical limitations that researchers and practitioners still wrestle with in large models.That said, the prescription is not a binary choice. The highest‑value profiles will fuse durable human skills with domain competence and technical literacy. Stakeholders—universities, employers, policymakers and technologists—should treat Garman’s advice as a strategic orientation: invest in human critical faculties while building tools, curricula and policies that help large populations acquire them. Companies must align their reskilling investments and workforce transition plans with the rhetoric, or risk eroding the very trust they need to navigate a decade of rapid change.
Garman’s short message to his children—“develop your critical thinking skills in college, no matter what subject you study”—is a practical, portable guideline for a volatile labor market. The work ahead is institutional: to make that advice actionable and equitable at scale, stakeholders must redesign learning pathways, measurement tools, and corporate transition policies so that people, not just machines, capture the gains of the AI era.
Source: Storyboard18 AWS Chief reveals list of skills other than coding that will define the AI age
