Why DragonflAI?
As some of you will know, we recently changed our name from CensorPic to DragonflAI, ushering a new version of our company.
History of CensorPic
I thought up the name CensorPic back in 2018 when forming the company, and distinctly remember throwing out some of the most unimaginative and derivative names at my friends until stumbling on CensorPic: a mash-up of Censorship and Pictures. In a rush to incorporate and really get started, I thought it was perfect.
Going forward, we raised money from private investors, won Scottish Edge under the name CensorPic, and were even awarded a trademark. But as we built our brand and connected with companies advisors globally, we kept meeting people who were somewhat uncomfortable with the connotation of Censorship being so intrinsically tied to our brand through the company name.
Censorship doesn’t evoke many positive feelings, especially in America, and due to excitement about creating what we thought was a clever name, we glossed over the negatives relating to a company almost called ‘censoring pictures’ that recognises image content. CensorPic was a recipe for branding disaster — it needed to change!
Brainstorming
Changing the name was quite difficult for me. Trademarking, buying domains and opening a freshly printed box of business cards with CensorPic stamped boldly on the top was the very manifestation of my desire to build a company. It had been the focus of my life for over a year at that point, but in January 2020 we went back to the drawing board.
The toughest question we had to answer in the name was “what do we do?”.
For us; off-line, nudity, protection, and real-time were some of the main summarised groups we chose to focus on building our new name from— messy on the surface, but it was a start.
We had everything from “Motti”, a Finnish military tactic, to “BodyBlock” which both looked great through the ‘Literal’ lens. However, I knew in my heart I couldn’t stand in front of people and pitch “BodyBlock” or “Motti” with the same conviction I had for CensorPic, so we had to go much more lateral in thinking.
An idea we kept coming back to was seeing an image in a camera lens but not actually creating it — exactly what we do as a company. This was then expanded to hunting where you can use a scope to see the target before making a shot. This was the light bulb moment.
Who were the best hunters?
Our incredibly accurate algorithms and rapid image detection times (sub 50ms) are fundamental to what we do; our safety tech needs to be correct and quick to ensure everyone using our software stays happy and safe. Looking at speed first, we thought about a famous spotted big cat of Africa.
But Cheetah is not only a predictable name, they are also only successful hunters 50% of the time — not good enough when our classification accuracy’s hover above 97%.
It wasn’t long before we discovered that natures most successful predator predicts how its prey will move, and adjusts its flight accordingly— the dragonfly.
It was an easy jump to combine our reliance on AI with dragonfly — it was perfect.
This was the name.
Logo Decisions
To drill down on what we do, we distilled two core things in terms of technical features of our product:
- Using AI
- Making unwanted content disappear (or not exist in the first place)
With these two features, we had the challenge of finding a designer who could marry up artificial intelligence and algorithms, with making things cease to exists — no mean feat for a non-designer like me.
Our 99 Designs competition gave us thirteen solid choices to choose from (after sifting out the numerous designers who seemed to copy a very popular node-logo from the first page of google images after typing “tech + dragonfly”). Some were flying off the screen while others were flat on the page. As inconsequential as that literal and figurative depiction of our new logo choices sounds, a flying dragonfly was apparently viewed as predatory, but that’s what we wanted to be, constantly on the hunt for indecent content.
However, I am big fan of community opinion so I sent out a poll to the residents of CodeBase to help inform my decision.
After getting back the, ahem… wrong poll results (in my opinion), I saw the top choice was a logo I hated. Although fantastically designed, it wasn’t for DragonflAI. I took this as a learning experience in being decisive for my company, even if others didn’t agree there are some decisions I just need to trust my gut on— so I picked my favourite.
There was only one thing off which CodeBase members were quick to spot — the body of the dragonfly looks like a ‘tadpole’ (or something slightly more ‘masculine’ to put it nicely), but we decided to roll with it at that moment as it fit with the type of content we moderate.
Side note: we didn’t want to go for the typical “tech company blue” so of course we ended up with the typical tech company blue…
It seems like moderating indecent content is black and white with the decision being either ‘safe for work’ or ‘NSFW’. In actuality there are many nuances to what could be considered acceptable depending on the situation. With our goal to redefine how moderation can be done, and provide the best possible online safety for children, the decision to colour our background grey seemed apt considering how we constantly improve our software to ensure what is acceptable can be viewed, and updating to keep up with changing thoughts on indecent content.
The final product can be described as follows:
- AI is the core of our company represented by the wings
- We endeavour to prohibit or delete illegal content represented how the wings are being ‘shaken off’ or ‘are disappearing’
- The flying dragonfly is hunting, representing our mission to develop the most accurate algorithms in the world that stop the proliferation of child abuse imagery.
The Tag Line
Unfortunately (or fortunately for a reader), there is nothing spectacular about how we came up with the tag line “Less Liability, More Protection”.
Simply put, we hope to bring less liability to platforms by keeping nudity, including child nudity, off social media. But also, since we can analyse and moderate content without it ever leaving the handset, we can offer much more privacy and protection to users.
In addition, given images never leave the handset, we are E2EE compatible, so users are able to have the benefits of encrypted messenger, and children cannot receive or send nude images.
This tag-line is still a work-in-progress and currently we are working with ‘Client-Side Moderation’ but DragonflAI is here to stay.
Our Company
This wasn’t meant to be a plug for our company, but we really do believe we offer the future of moderation, and given our recent and negotiated full approval of the DragonflAI trademark, we are thrilled to be able to tell you about our journey to re-brand.
Our algorithms run entirely on the device, detecting content both in real-time and offline meaning we can prevent nudity being uploaded to your platforms, or being live streamed to fans.
If you are interested in joining our beta testing then email Hannah@dragonflai.co and I will get in touch with you.
