(This column originally appeared in Forbes)
In 2007 the iPhone’s introduction spawned an avalanche of mobile apps. When OpenAI introduced ChatGPT 3.5 to the public in 2022, the same thing happened: a proliferation of apps, tools and gadgets that use generative AI to accomplish certain business tasks.
Most of the mobile apps from the iPhone launch era are no longer around. It will be the same with today’s AI apps. That’s because these things usually take one of four roads.
A very, very, very few will actually scale into something big and become a genuine platform. Think Slack or WhatsApp or Square, which all started out as much smaller mobile applications and evolved. Some AI apps will get purchased by a bigger tech company and rolled into their offerings. Other larger software companies will simply write their own AI apps that replace the existing ones. A handful of these apps will stick around but will recede into the shadows as a niche tool with a small fanbase.
You can argue with me. But in the end 99% of the AI applications you hear about today won’t be around in just a few years. It’s not that they’re not good. It’s just Darwinism. Here are eight examples.
This is a great application that uses AI to monitor meetings, transcribes conversations, summarizes discussions and creates actions. Does anyone doubt that Microsoft Teams, Zoom or any other more dominant meeting platform will offer these same features? Most already do or are rolling out these features. Why get Otter when your existing meeting application already provides this functionality?
There are countless AI apps that promise to take a standard photo of any schlub like me and turn them into Brad Pitt (or at least a better version of themselves). Sounds impossible — particularly in my case — but it happens. I point out Profile Pic Maker because it’s one of the best of them. But c’mon — like this won’t be built into the next camera app on your device? Of course it will.
This is another great AI platform that can take long videos from YouTube and summarize them, saving a user hours of watching time. So why doesn’t YouTube just do this? Maybe it’s because the longer someone watches a video, the more ads they watch, so that’s a disincentive. Until YouTube makes this part of their paid offering or YouTube TV. Sorry, but I don’t see this type of technology as a stand-alone long-term play. Perhaps as a niche survivor.
I love this application. It uses an AI assistant to schedule meetings, navigate conflicts and create agendas. Among other powerful features, the platform literally lets users do this just by chatting with it or sending a message through Slack and then it finds the time on everyone’s calendars. There’s no question that this is a problem needing to be solved and Xembly does a good job solving it. Until Microsoft, Google, Slack and all the other larger customer relationship management (CRM) and workspace app providers just build this into their own platforms — or buy Xembly.
This app converts videos into 130+ languages in a matter of minutes. Very cool. Will YouTube offer this feature? Hmm…130 languages plus more viewers in more countries equals more eyeballs and…ka-ching…more advertising dollars. Seems like a no-brainer. Or otherwise, will a larger video production platform — say Adobe — do the same? Of course they will. Even so, this could still survive as a niche product for video people who are a fan of a certain type of production software which doesn’t offer this functionality out of the box.
Because I’m in the CRM world I’ve played around with Lavender and it’s fun. The software will “coach” you to write better emails. It’ll score the emails you write and offer improvements and personalization suggestions. “Become an email wizard!” the company promises. “Level up your team! Make email magic happen!” As if Salesforce, Zoho, and every other major CRM player won’t just do the same? It’s only a matter of time.
Want to create a large language model for all of your company’s documents and files? This way you can train your LLM so that your employees can ask it anything about your policies, procedures and rules and get accurate and consistent results. That’s what AskJack will do and this can apply to human resources, operations, help desk, customer service and legal departments. But wait…both Microsoft and Google are already doing this by creating LLMs out of a company’s emails, calendars, OneDrive and Google Drive documents. Within two years any Office or Workspace user will have the same functionality that AskJack offers.
Here’s a specialized LLM for contracts. The AI application bills itself as a “negotiation partner” for both lawyers and non-lawyers that will help identify risks, explain legal jargon, make tailored suggestions and give other insights to any legal documents or contract you allow it to review. Assuming the output is reliable (I admit to not testing it yet) I wonder long term who will use this tool? Won’t this kind of thing be included in a mainstream law firm platform like Clio or Smokeball or be made available to consumers via bigger players like LegalZoom and LegalShield? Probably so.
Agree? Disagree? To be clear, these are great apps and I highly respect the people that developed them. I’m also assuming that a few of the app makers aren’t looking to be the next Snapchat — they’re doing what they’re doing to create new tech and make a few short term bucks and good for them. But for the rest, I’m just hoping they’ve got a good exit strategy. Because things in the AI world are moving fast and most won’t last on their own for very long.
