Nobody voted for distribution

Articulate held its keynote in Orlando last week. Before it, Tim Slade ran a poll asking what people actually wanted from the future of workplace training. Around 40 percent said major Storyline updates. 26 percent said Rise updates. 32 percent said something entirely new.

About 1 percent picked improvements to how courses get distributed.

The keynote has now happened, and that 1 percent is still the number I cannot stop looking at.

What was announced

Three things. Articulate 360 AI, which is the authoring suite most L&D teams already own, with generation, avatar video and translation into 80-plus languages built in. Frontline, a new product for people outside L&D who need to turn a document into a "training kit" and share it by link. And Nova, the agent underneath both, which Articulate says is "backed by 20+ years of learning science".

All of it is metered by credits. Every account gets 20,000 per seat to start. Avatar video costs a thousand credits a minute. Credits expire at the end of the term, and when they run out the features that use them stop until someone buys more.

The line from the stage that I wrote down was Brian Gil's: "The constraint has never been your expertise. The constraint has been how far that expertise can reach."

That is a good line. It is also, precisely, the answer to the question the 40 percent asked. A better way to make the thing, reaching more people. It is not a question about whether the thing needs making.

Tim's thesis, and where I land on it

Tim's essay before the keynote argued that we have spent two decades trying to escape a model nobody really wanted. Build something in advance, park it somewhere, ask people to go and complete it. Every wave of technology promised to break that and then did not. The LXP gave us better thumbnails. Microlearning made the static content shorter. xAPI let us measure more of it. His phrase for the architecture underneath is a "21st-century vision trapped inside late-1990s technical architecture", and his name for what comes next is the post-course future. Not that courses disappear. That "the course stops being the automatic answer".

I think he is right, and I think the keynote proved him right in an unexpected way. Frontline is the first product from a major vendor whose output is not a course. It is a kit: a guide, a quick reference, a scenario, a deck, shared by link, no SCORM. That is a post-course product, from the company that owns the course. The 32 percent who wanted something new got it. They may not have noticed.

Now the pushback, which is the same one I made before the keynote and which nothing on that stage changed.

The auditor needs an object

I sell into banks, telcos and government groups in Saudi, the UAE, Uzbekistan and Tajikistan. From here the post-course future looks further away than it does from Phoenix, and for a reason that has nothing to do with pedagogy.

In these markets a regulator requires evidence that a named person was trained on a named thing by a named date. Evidence needs an object. A completion record. A certificate. Something you can hand an auditor. You cannot hand an auditor a conversation that happened in Teams, and you cannot hand them a kit shared by link either.

So the course survives here not because anyone thinks it is the best way to learn, but because it is the only artefact the compliance system knows how to read. That is a much stickier reason than habit. Tim exempts compliance and certification in his essay, which is right. I would just say the exemption is bigger than it looks. In my part of the world it is most of the budget.

What the comments got right

The replies under Tim's video were more useful than most analyst notes. A few things came up again and again, and they are the things I would build a company around if I were starting today, which I suppose I am.

Generation is the floor now. Nobody in that thread was impressed that a tool can produce a course. Every tool can. Several people said they already do it in a general chat assistant and paste the result into whatever they have. The interesting question is what happens on the tenth course, and the fortieth, when three different people have built them and none of them match.

Consistency to a standard is the thing people actually miss. Not speed. The word "brand" came up more than the word "AI". Instructional designers know what a good course looks like; what they cannot do is make forty of them look like each other when the business wants them by Friday.

Credits frighten people. The most upvoted worry was not about quality. It was about a bill nobody can forecast, in a tool where the meter runs on video minutes and per-character translation and the feature switches off when the pool is empty.

Translation is not the same as writing in a language. Eighty languages sounds like a lot until your Arabic reviewer sends the module back because the register is wrong and the layout broke on the quiz.

The part that implicates me

I run a company that produces course content. If the thesis is right, I sell the packaging. Faster and cheaper packaging, in languages the agencies handle badly, but packaging.

I am fine with that, and not because I am hedging. The demand is real and it is now. A bank with eleven thousand staff and a regulatory deadline this quarter does not care what the category looks like in 2030. And the work of getting content out of documents and into something a named person can approve and an LMS can record is exactly the work you have to be good at before you can do anything more interesting. You do not skip it.

But I would rather be honest about which part is the destination and which part is the road. What we build now is the road: the same standard on the fortieth course as the first, Arabic that was never English, a check of every screen against the source, an approval with a name on it, one package in the LMS you already run, and a bill you read before you built. Arvand Bank, a thousand people, mandatory training every six months, runs on it today. That is the boring, defensible part.

The destination is the 1 percent. Training that reaches a person when the policy changes rather than when the calendar says so, with a record the auditor can still read. Nobody voted for it. It is still the only thing on the ballot that changes the architecture.

Prompted by Tim Slade's essay "Articuland 2026: Is Articulate Storyline Still the Future?", published in The eLearning Designer's Academy the day before the keynote. The poll figures and the post-course framing are his. Product details are from Articulate's own announcement and support pages as of 15 September 2026.