I’m disappointed by the ending of PullString Author as a product. It was a unique program in the chatbot landscape. Author was usable by non-programers and testable without installing anything additional. Their blog post on their “Voice-First Future“
Does anyone know of a similarly powerful GUI application? The others I’ve tried were laughably clumsy and had 1% of the power.
At the bank where she works, my wife and another analysts won an internal innovation contest with it and are pitching to senior management soon. Author makes it possible to convince management to try the project because the investment is so much lower compared to coding from scratch. A company can dedicate a few people for a month and have a powerful chatbot to explore their use. But if they have to dedicate analysts and a software team for several months, that’s much less likely to happen. The potential outcome is the same, but the investment is very different.
Another example: I work for a healthcare software company. We are considering text chatbots but not voice, because of data security problems – not knowing who is in the room and listening, not trusting voice authentication, and how painfully inexact voice instructions are. Text-based chat is more secure and is less disruptive in user’s lives (notification, asynchronous ongoing interactions). And with its visual dimension it’s much more powerful for communication. Voice, in our analysis, is only useful for trivial one-shot tasks of a few sorts. It’s an afterthought.
Text conversation is email, SMS, Facebook, Chat. Voice is… the _last way_ most of us communicate. On Reddit, Facebook, and email we are not leaving voice recordings of our thoughts, because that’s the slowest way to interact and accomplish anything. It’s far more powerful than voice. Voice interaction is an affordance for when our hands are busy.
This seems like a poor move to me and aiming really far at the future when there text-based chatbots are just now starting to enter mainstream awareness.
Author was for building conversations. I don’t know why PullString thinks preferring voice is a superset of that, rather than a massive narrowing. I can’t follow the logic of throwing out the highly-developed conversation designer just because the end interaction is voice instead of voice _or_ text.
I thought PullString was positioned to be a very dominant player at the enterprise level. They have a product that visually makes more sense than any other – you can see conversational structure. Non-experts can change words and add phrases. It creates ownership by giving power to non-programmers. I’m surprised they disagree and are withdrawing from their strong position.
I do hope PullString comes out with something great – I have benefited from the free Author and enjoyed it immensely, though I haven’t launched anything commercial with it yet. They must be on to something big to make such a drastic change.