A Chicken In Every Bot
Bot hype battles bot reality
After a (purposefully) vague tweet last week, Re/code reached out to get my take on bots, for an overview post they were writing. The quote they used is as follows:
“To be honest, I’m a little worried about the bot hype overtaking the bot reality,” said M.G. Siegler, a partner with GV, the investment firm formerly known as Google Ventures. “Yes, the high level promise of what bots can offer is great. But this isn’t going to happen overnight. And it’s going to take a lot of experimentation and likely bot failure before we get there.”
I suppose this paints me as a bit of a “bot bear” in an age of “bot bulls,” leading up to Facebook’s f8 conference keynote tomorrow, where the company is expected to unveil some sort of new bot platform/tools. So I thought it was worth elaborating a bit — I’ve been thinking about this for a while. Below are my full thoughts I sent over to Re/code:
Yes, the rise of the bots has been fast and furious in startup land.
I think there are two main factors converging that are leading to this. One is the increasing importance of messaging services and their extension into being platforms. This is both on the consumer side with services like Facebook Messenger and on the enterprise side with services like Slack. The second is the inundation of apps. Both the app stores and peoples’ phones are now saturated and we’re finally at the realization that not everything needs (or can sustain) a full, native app.
The third element, which I don’t think is actually quite here yet, is the AI/machine learning part of the equation. While we’re starting to see this with the big companies, with most startups, a lot of the bot functionality is still actually manual, done by people (or at least, heavily assisted by people), behind the scenes. This will change over time and make all these bots more powerful, of course.
To be honest, I’m a little worried about the bot hype overtaking the bot reality, already. Yes, the high level promise of what bots can offer is great. But this isn’t going to happen overnight. And it’s going to take a lot of experimentation and likely bot failure before we get there. I’m not sure people are being appropriately primed for that right now.
The other side of this is that because it’s relatively easy to get a bot up and running — more so than a native app, or even a website in many cases — there’s some concern about the same inundation problem happening. And happening even faster. Just as everything probably doesn’t need an app, does everything need a bot? Probably not. So, as ridiculous as it sounds, we may see a “bot backlash” sooner than you think.
But this will all naturally shake itself out. And I do think that in the long run, there will be a range of interesting startups/services built around bots. But it’s going to take some time, and there will still be bumps along the way. Despite the hype, it’s super early.
So there you go. I’m hardly in the “Bots Suck” camp, but I’m not exactly a bull either. It should probably be the least controversial thing in the world to suggest that it will take a while for bots to live up to their potential. But the amount of buzz in recent months — and especially the last several weeks — around bots has me a little worried that expectations are being set far ahead of reality here.
And that’s a double-edged sword. On one hand, it has already super-charged the interest and excitement around bots on the entrepreneurial side. But the flip-side is that if we’re suddenly all flooded by bots, and in particular, low-quality bots, it could harm a nascent ecosystem.
Said another way: it sure feels like the Fart Bots are coming.
Of course, the Fart Apps of yesteryear didn’t ruin the app ecosystems. But as noted above, apps — even stupid fart apps — have a much higher barrier to entry versus crappy bots. Not only did you have to write (at least some) native code, you had to get them past Apple’s App Store gatekeepers. With bots, they’re both easier to write and easier to deploy — and because so many bot platforms are already rising — presumably much easier to deploy.
Now, a well-done bot will probably be harder to do in many respects than an app because of the immense back-end magic that will have to happen. But again, I just don’t think we’re anywhere near where we need to be with regard to that back-end magic, so it’s going to be awhile.
My advice to anyone building a bot right now is to keep it insanely simple and focused. Do one thing great and let the technology catch up to the hype. If you try to fake the magic at a massive scale, you will likely fail. Baby steps.
