AI has made coding faster, but faster coding does not automatically mean faster work.

If your day is already packed with meetings, decisions, reviews, debugging, planning, and problem-solving, generating code in half the time only solves one part of the workload. This is where the gap between AI’s potential and real-world productivity becomes important.

The popular expectation is simple: give someone AI, let the machine write most of the code, and productivity should explode. But software development has never been just about typing code. Much of the difficult work happens before a developer writes the first line and after the code has been generated.

That distinction matters if you are trying to use AI without creating unrealistic expectations for yourself or your team.

Consider a senior developer working an eight-hour day. Before AI, roughly 1.5 hours might go toward writing new code, while another 1.5 hours goes toward reading and debugging. The rest can involve architecture, reviews, documentation, testing, mentoring, deployment, and meetings.

Home furniture brand Blu Dot blew up on CTV with help from Roku Ads Manager. Here’s how:

After a test campaign reached 211,000 households and achieved 1,010% ROAS, the brand went all in to promote its annual sales event. It removed age and income constraints to expand reach and shifted budget to custom audiences and retargeting, where intent was strongest.

The results speak for themselves. As Blu Dot increased their investment by 10x, ROAS jumped to 2,308% and more page-view conversions surpassed 50,000.

“For CTV campaigns, Roku has been a top performer,” said Claire Folkestad, Paid Media Strategist, Blu Dot. “Comping to our other platforms, we have seen really strong ROAS… and highly efficient CPMs, lower than any other CTV partner we've worked with.”

Using Roku Ads Manager, the campaign moved from a pilot to a permanent performance engine for the brand.

Learn more

If AI makes coding three times faster, writing new code could fall from 1.5 hours to about 30 minutes. That sounds impressive, but the rest of the job does not disappear.

The result in the example is a reduction from eight hours to approximately 6.75 hours, or about a 15% productivity improvement.

That is useful, but it is nowhere near the mythical 10x transformation.

The reason is straightforward: AI is extremely effective at certain tasks, while other parts of the job depend heavily on judgment, context, communication, and decision-making.

Tip: Measure AI against the entire workflow, not just the time spent generating code.

One of the easiest mistakes is assuming that software development is primarily a typing exercise.

It is not.

A developer still needs to understand what problem is actually being solved, determine how the system should behave, identify constraints, evaluate trade-offs, understand existing architecture, and turn vague requirements into something that can actually be built.

AI can help with pieces of that process, but it does not automatically eliminate the need for human reasoning.

This becomes especially important for senior engineers. Their value often comes from knowing what should be built, why it should be built, and how it should fit into a larger system.

That is why someone who writes excellent code but cannot reason about systems, collaborate with others, or translate ambiguous requirements into concrete work is still missing some of the most important skills of the role.

Being able to produce code is increasingly becoming table stakes. Knowing what code should exist in the first place is much harder to automate.

There is another uncomfortable part of the productivity conversation: AI-generated work is not automatically efficient to consume.

A long AI-generated requirements document may contain plenty of technically correct information while still making the important decision harder to find. The same can happen with tickets, documentation, explanations, and proposals.

The problem is not that AI produced something useless. The problem is that more content does not necessarily create more clarity.

If someone spends less time producing a document but everyone else spends more time understanding it, the organization has not necessarily become more productive.

This is why AI adoption needs to consider the entire flow of work rather than celebrating how quickly one person can produce an output.

Tip: Ask whether AI is reducing total effort across the team, not simply reducing the effort of the person using it.

There is an interesting twist in the numbers.

A junior developer in the example spends more of the workday actually writing code. Before AI, coding accounts for about 2.75 hours of an eight-hour day. With AI assistance, that falls to roughly one hour.

That produces a total workday of about six hours, representing a 25% improvement.

The junior therefore receives a larger productivity boost than the senior developer in this scenario.

That challenges the assumption that AI makes junior developers unnecessary.

In fact, a junior who uses AI properly can potentially accelerate learning while receiving assistance with repetitive implementation work. The important distinction is whether AI is being used to understand the work or simply avoid learning it.

A developer who asks why a solution works, compares alternatives, tests assumptions, and uses AI to explore unfamiliar concepts can build capability much faster than someone who simply accepts generated code.

AI becomes far more valuable when it acts as a learning partner rather than an automatic code vending machine.

Tip: Use AI to explain, challenge, and teach—not just to produce the final answer.

TLDR Marketing is the free daily brief that 330K+ growth marketers, performance marketers, and CMOs actually read. The most interesting stories in marketing, curated and summarized in 5 minutes.

Subscribe for free

The most important shift may be that AI is changing which parts of technical work deserve the most attention.

When code generation becomes faster, the bottleneck moves elsewhere.

Architecture becomes more important. Requirements become more important. Testing becomes more important. Reviewing AI-generated work becomes more important. Understanding systems and making good decisions becomes more important.

In other words, AI may reduce the value of typing speed while increasing the value of judgment.

That creates a different kind of productivity advantage. The person who can clearly define a problem, give AI the right context, evaluate its output, and quickly identify what is wrong can accomplish far more than someone who simply knows how to generate code quickly.

The machine can accelerate execution, but someone still needs to determine whether the execution is heading in the right direction.

The smartest expectation is not that AI will suddenly eliminate most of the work.

It is that AI will gradually remove friction from specific parts of the work.

Today, that might mean generating boilerplate, explaining unfamiliar code, creating tests, debugging simpler problems, or helping a junior developer understand a concept. Tomorrow, AI may become capable of handling much larger portions of architecture, planning, and operational work.

But until those capabilities become reliable enough to trust, productivity gains will remain uneven.

For someone already overwhelmed by a full calendar, the goal should not be to fill the newly created time with even more tasks. The real opportunity is to use AI to create breathing room for the work that still requires human judgment.

That might mean spending less time writing repetitive code and more time thinking through a difficult design. It might mean reducing debugging time so there is more room for mentoring. Or it might simply mean ending the day with fewer unfinished tasks.

Tip: Treat every hour saved by AI as capacity you can deliberately redirect, rather than another hour that must be filled.

The promise of AI productivity is not necessarily that one developer suddenly becomes ten times faster.

The more realistic transformation is that different parts of a workflow become incrementally faster, while the bottlenecks move to areas AI has not yet mastered.

A 15% improvement for a senior engineer may sound modest compared with the headlines, but repeated across hundreds of people and sustained over time, even modest improvements can matter.

And as AI becomes better at planning, reasoning, testing, documentation, and other parts of the development lifecycle, the productivity equation can change again.

For now, the lesson is simpler.

AI can make people faster without making the entire organization fast.

The biggest advantage will belong to those who understand where the real bottlenecks are, use AI where it genuinely helps, and continue developing the human skills that become more valuable as machines handle more execution.

The future of productivity is not about becoming more machine-like.

It is about becoming better at deciding what deserves your attention in the first place.

What’s your next spark? A new platform engineering skill? A bold pitch? A team ready to rise? Share your ideas or challenges at Tiny Big Spark. Let’s build your pyramid—together.

That’s it!

Keep innovating and stay inspired!

If you think your colleagues and friends would find this content valuable, we’d love it if you shared our newsletter with them!

As the world becomes increasingly digital, this question will be on the minds of millions of people seeking new income streams in 2026.

The answer is—Absolutely!

Thank you for taking the time to read today’s email! Your support allows me to send out this newsletter for free every day. 

 What do you think for today’s episode? Please provide your feedback in the poll below.

Share the newsletter with your friends and colleagues if you find it valuable.

Disclaimer: The "Tiny Big Spark" newsletter is for informational and educational purposes only, not a substitute for professional advice, including financial, legal, medical, or technical. We strive for accuracy but make no guarantees about the completeness or reliability of the information provided. Any reliance on this information is at your own risk. The views expressed are those of the authors and do not reflect any organization's official position. This newsletter may link to external sites we don't control; we do not endorse their content. We are not liable for any losses or damages from using this information.