Talking Moose: A Mac Original That People Still Write About

In 1986, a black-and-white moose appeared on Macintosh screens and told jokes aloud, his mouth moving in time with the words. He was made by a Canadian physician, Dr. Steven B. Halls. The code was his own work, written in assembly language.

The character grew familiar to Macintosh owners throughout North America and Europe. Many people knew the Talking Moose. A few still write to its creator, wondering whether a new version is coming.

It is a long afterlife for a desktop diversion.

Antlers, pixels and a mouth that talks

A realistic human face was never needed for the original Talking Moose. On a small, black-and-white screen, his antlers made him recognizable. Halls liked moose, and the shape worked well with so few pixels.

Making him speak convincingly was a harder problem. The mouth needed to follow the speech, not simply open and shut while a voice played.

Halls found a way by watching Macintalk, the Macintosh speech engine, in memory. It let the Moose determine the sound being played and choose a matching lip shape. The technique worked because early Macintosh software could inspect memory across applications much more freely than modern systems allow.

It was a watcher, not a modifier of the speech engine. Code written with care turned speech information into visible movement, so a synthetic voice had a face.

The original character could also comment on window and document names in familiar programs.

A magazine award and an unusual greeting

By around 1989, the Moose had gained color. Halls's pictures of the color version include scans of an image produced with an early Macintosh color printer. You can still make out the printing dots.

It was more than users enjoying the jokes. The software won MacUser magazine's "MacUser Editors Choice Award" for "Best Desktop Diversion" in 1991.

There was also a reception Halls has never forgotten. At an Apple Developers conference during those early Macintosh years, fellow developers greeted him as "a God". They believed he had coded something impossible.

It was their reaction, and it speaks to the surprise of watching the Moose work. Today, a face matching a computer-generated voice may seem ordinary. On the early Macintosh, Halls made it happen with assembly code and an understanding of the speech engine's memory.

The award and the developers' greeting show two sides of the same program: users enjoyed having it around, and programmers valued the work behind it.

Career, children and the requests that kept coming

In later years, Halls shifted his attention toward his radiology career and family life with children. The Moose was not at the center of all he did.

But requests for a new version kept arriving sporadically. They continue even in 2026.

Software can be well known while people are using it. It is quite another for someone to remember it years later and go to the trouble of contacting its author. The requests have kept the old character in Halls's life, long after those first Macintosh screens.

A new kind of Moose on Windows

The later Windows project gave the character a head in three dimensions. Its ears could move, and its eyes could look around. For lip synchronization, it used timing information from Microsoft's Speech API, unlike the Macintalk memory method.

Movement was different too. The Moose used procedural animation instead of replaying identical clips. Where a movement ended and how quickly it happened were varied by algorithms. Pauses varied as well, and smooth acceleration and deceleration made motions look natural.

What the character could say grew through plugins. Alongside jokes came news and weather, and reminders and notifications gave more practical reasons to hear from him.

In a June 2016 development update, Halls described a working Pushover and IFTTT connection. The Moose could announce approaching rain, or say "Welcome home" when a phone entered a home geofence. By 2019, alpha testers were describing their experiences with the Windows software, with installer instructions in four languages.

None of this turns the original achievement into a story about conversational AI. The Macintosh Moose spoke through programmed behavior and speech synthesis. Halls had worked out how to match the mouth to the sounds, and so it moved.

A magazine called it the best desktop diversion. Fellow developers thought the coding was impossible. Requests for another Moose still reach the physician who proved it could be done, decades later.