AI can make parts of data science faster, but Yu Dong, the article’s author, argues that its bigger effect is letting one data scientist take on a wider range of work. In Dong’s account, the role shifts from writing every query and script toward planning, reviewing, and deciding whether an analysis is trustworthy. That is a personal perspective—not a measured finding about the profession as a whole.
Is AI changing what data scientists do, or just helping them do it faster?
Dong’s answer is both. Generating code can reduce hands-on execution, while AI-assisted workflows can also bring research gathering, analysis, engineering, and stakeholder communication into one person’s remit. As Dong puts it, “AI doesn’t just make the same DS job faster. It changes what one data scientist can reasonably own.”
Dong says that over the preceding six months, they had rarely written SQL or Python manually, relying on AI to generate much of the analysis code and taking a more active reviewer role. That timeframe describes Dong’s own workflow; it is not a statistic about how often data scientists generally use AI or stop writing code.
Which parts of the work can AI help cover?
Research and analysis planning
Dong describes using tools to gather discussions and previous research, then plan an analysis. The potential change is not limited to getting help with a query: AI can assist earlier, in assembling context and structuring the work to be done.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errors#1 Best Overall
Code and execution
AI-generated SQL and Python can reduce manual coding in Dong’s workflow. The human role does not disappear in this account; it moves toward specifying the task, examining what the system produced, and judging whether the result is sound.
Engineering and model work
Dong also recounts using AI coding tools to help with data engineering and model work. Broader execution capacity still leaves consequential choices to people, including selecting an appropriate data model and reviewing changes before they reach production.
Stakeholder-facing communication
Dong describes AI helping draft a write-up for stakeholders after an analysis. This extends assistance from producing technical work to explaining it, but a draft still needs a human to check that its claims match the analysis and make sense in the business context.
Why does human review still matter?
More output is not the same as more reliable output. Dong highlights two risks: people may distrust incorrect AI-produced results, or trust incorrect results too readily. Both can undermine decisions, though in different ways.
Business definitions and semantic layers—the shared meanings and rules behind metrics—also need ongoing maintenance and human review. If those definitions are stale or wrong, faster analysis can simply produce answers that are consistently misleading. In Dong’s view, the ability to generate code increases the importance of checking inputs, definitions, and results rather than removing that responsibility.
Does broader ownership also mean more workload?
Dong reports supervising several projects run in parallel with agents. That can expand how much work one person coordinates, but it also creates more context switching: keeping track of separate tasks, decisions, and outputs. Dong describes the attention burden and the possibility of higher delivery expectations alongside the added capacity.
Rank #4
This is a concern from one person’s account, not evidence that using agents causes burnout across the workforce. It does, however, point to a practical trade-off: doing more kinds of work at once may increase coordination demands even when individual tasks take less manual effort.
What does this mean for a data science career?
Dong’s view is that writing code may become less distinctive on its own as AI makes routine execution easier. Technical judgment, business understanding, selecting the right question, and checking AI outputs may matter more because they determine whether the work is useful and trustworthy.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
The shift raises an unresolved question about junior staff. If some routine execution changes, early-career workers may have fewer opportunities to build skills through those tasks. Dong flags this concern but does not establish how training paths will change or offer a settled solution.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What the account does—and does not—show
Dong’s argument is that AI expands the scope a data scientist can cover, rather than simply making an unchanged job faster. The examples span research, analysis, coding, engineering, and reporting, while preserving a need for human judgment and governance.
The account is a personal report, not a workforce study. It supplies no survey, controlled experiment, sample size, or labor-market data to show how common these changes are or what they mean for employment overall. Dong’s concluding phrase, “AI is not shrinking the DS job. It is stretching it,” is best read as the author’s framing of that experience—not a proven forecast for every data scientist.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Free tools Windows power users keep installed
One-click scans. No signup required.
