I appreciate your response.
My overall view is ML/AI is still in its infancy and there is still a lot of growth to be had, so there will still be a significant appetite for data for quite a long time to come. Appen has now reshaped their business to respond to customers with a variety of needs and price points in order to capture a greater market share.
While there are several competitors to Appen, none of them are making a profit. Scale AI does a good job automating most of their tasks but using AI to annotate data has its limitations. Even neural networks seem to struggle when applied to real world scenarios. Autonomy is at least 15-20 years away, and while SSL is currently a potential threat to Appen it will simply just acquire an SSL platform if need be. You must remember that Appen is not a static company, it is a market leader. Market leaders always respond to new developments because they can, while other companies struggle to survive.
In my view, there is not yet a substitute for good quality annotated data. Sure, in the future there may be, but none have yet passed the “real world” test. All of these other tools and techniques are mostly in the research and development stage. It’s kind of like saying Pfizer should be concerned of this small biotech in Australia. Don’t get me wrong, they are great ideas, but that’s all they are at the moment, just ideas. I wouldn’t even bother focusing on transferred learning, SSL or unsupervised learning until they become a threat. Then guess what, Appen will just acquire the platform – simple.
Since you responded to some of my discussion points, I wanted to share with you some of my views in regards to China, Govt and Regulations (three major future catalysts for Appen):
China
China are renowned for cheap, low quality products ranging from manufacturing products through to construction. This is a tag that has stuck with them even to this day, but it is one that they are desperate to shake off. This is because China’s standards of living and wages have improved dramatically over the years, so the only way they can continue to compete is to charge more for their products. Now throw in inflation, political and trade issues, and China has a recipe for disaster. Soon China will have to bite the bullet and start producing higher quality products, otherwise people will just start buying from elsewhere. This includes AI products.
Now considering China wants to be a tech giant (notwithstanding the semiconductor and security issues), the only way they are going to produce high quality AI products is to use high quality annotated data. As of today, AI products from China are substandard.
Regarding your reverse engineering comments, China’s IP theft is normally strategic. If they wanted to copy Appen’s business model they would have done so already.
Govt
Appen are aiming to win contracts with US government agencies and currently have an office in Washington DC specifically for this purpose. They also have some of the major US tech giants as customers, who are tightly aligned with national security agencies. IMO it won’t be difficult for Appen to get a foot in the door with US government agencies, and if they do, this will easily open the door to contracts with other govt agencies from allied countries.
Regulations
It is my view that any form of regulation will help (not hinder) Appen. This is because those same regulations will also apply to its competitors and will increase barriers to entry. Appen is already an established market leader and the only data annotation company making a profit. They are able to rapidly adjust their operations to change to the changing market conditions, so regulations shouldn’t impact Appen severely.
There are currently no standards for ML and AI, but countries are actively working towards developing standards for AI, and policy frameworks, as important national priorities. This includes ‘responsible AI’. Addressing bias in AI will also be part of these standards, both of which Appen has incorporated into their business model.
In terms of slavery, I take it your referring to Ethical Sourcing and Modern Slavery Policy. This is something that is impacting several businesses worldwide and Appen has addressed this here:
https://appen.com/global-ethical-sourcing-and-modern-slavery-policy/
Whilst this is a business risk, Appen has taken steps to address it. For example, Appen has automated some of the data annotation processes, thus increasing their margins, which can potentially be used in the unlikely scenario to absorb any negative consequences resulting from this.
Expand
(20min delay)
