Is dictation becoming more popular?

Yes, among developers and founders. The word "dictation" appeared in 682 Hacker News stories and comments in the 12 months to September 2026 ¹. That is more than double the count for the twelve months before. "Speech to text" appeared 616 times ², up by a smaller margin, and "speech recognition" 249 times ³, which is nearly unchanged.

The data comes from 23 data points ⁴ that count stories and comments containing each phrase, per calendar month since January 2015. It measures talk, not use.

Mentions of dictation grew from a handful to dozens a month

In January 2015, "dictation" appeared 4 times in a month. In September 2026 it appeared 80 times. The chart shows the monthly path for dictation and speech to text.

"Speech to text" went from 2 mentions in January 2015 to 52 in September 2026. "Speech recognition" went from 12 to 28 over the same span. Its starting point was the highest of the three, which tells you the older, more technical phrase once dominated the conversation.

The vocabulary has shifted toward everyday words

The three phrases tell different stories:

• Dictation is the phrase for the act of speaking to write. Its growth suggests people now discuss using it, not only building it.
• Speech to text is a product-style phrase, common in app names and feature lists.
• Speech recognition is the research and engineering term. Its flat count suggests the technical conversation has not expanded the way the practical one has.

Here are the rows for these terms, with the prior twelve months beside them.

"Transcription" is the biggest of the general terms, with 1,627 mentions ⁵. It also grew over the prior year. Part of this is the use of speech models for meetings, interviews and video, which is a different job from dictating your own text.

Newer product names explain much of the rise

Mentions of specific products grew quickly as well. "Parakeet", a speech model, had 434 mentions ⁶, and "wispr flow", a dictation app, had 137 ⁷. Both grew sharply against the year before. As more tools launch, more people write about dictating, which feeds the general terms above. For a product-by-product view, see which dictation apps developers talk about most.

What these counts do and do not measure

A phrase search counts any mention, including criticism, and nothing is filtered by meaning. A post saying "dictation never works for me" adds to the total just as a recommendation does. Some words have other meanings: "parakeet", "moonshine" and "whisper" are also a bird, a drink and a quiet voice. One long thread can add hundreds of mentions in a single month, so a single month is a weak signal and the twelve-month totals are more reliable.

The data also covers only one community. Hacker News readers skew toward software developers and founders. It cannot tell you how many people dictate, how often, or whether they are satisfied. For that you need usage surveys and market data, not forum counts. The figures here show that attention is rising, which usually follows better tools, but attention is not adoption.

What this means if you are deciding whether to try dictation

The tools people discuss today are different from those of a few years ago. Many now run on your device and use modern speech models, so accuracy and speed are better than the older tools many people still remember. If you tried dictation years ago and gave up, a fresh trial is reasonable.

Start with the free option built into your computer or phone, and move to a dedicated app only if you hit a limit such as poor accuracy, missing commands or privacy concerns. Our guide to AI dictation apps compares the options, and the market overview puts the products side by side.

Keep reading

• The full dataset shows every row, its history and its caveats.
• Dictation software is our overview of the whole subject.
• Speech recognition explains how the technology works.