Web Summit inspirations: how #AI can bring opportunities and not take away jobs… just yet
Originally published November 16, 2017
Two months on the road: United Nations General Assembly, Techfugees and Web Summit. From global challenges that still appear impossible to grasp to the massive fear of soon ‘fully automated world’, thanks to artificial intelligence. Indeed, Web Summit was a transformative event for me as a former UN worker: Stephen Hawking’s warning, AI Alexa ordering your coffee, Uber ’s flying cars, hundred of fin-tech solutions to ease how we do transactions and talks on what to do when robots will do work instead of us.
Now I am back to reality, Lebanon is a place where:
- financial inclusion is yet to reach at least 50%.
- there is no public transport of any sort.
- there is a garbage crisis.
- 80% of people work in the informal market mostly in the sell or trade and becoming a services-oriented economy is still a dream.
- Sectarian politics are present and appear medieval from an outsider’s point of view.
- 2 million refugees with less than 1% have a work permit.
This is when you realize how much inequality we still have in this world, and having checks and balances between these two extremes is necessary. We are yet far away from the perfect world.
Now how we can bring digital revolution where it is the most needed? This is what I learnt at the Web Summit…
We are only in the beginning of the data revolution. On average each person will generate 1.5 GB of data every day by 2020. Now it is only 600–700 MB daily, so eventually it will double. That means more and more data will be digitized and can be used for variety of automated processes — including autonomous driving. Just for the reference to power the ability of autonomous car to drive it is needed 4000 GB of data. Also, according to the IDC (International Data Center), only 15% of businesses in the Middle East have the online presence. (Source: Intel )
There is a big market need for data digitization and labeling that is yet to be fulfilled!
Some would argue that now everything is open-sourced, more data sets are being labeled and stored at Kaggle and there are depositories that make use of this data such as IBM Cognitive Business Whatson, AWS Startups use Alexa and image recognition technology to power AI businesses. Google Cloud just recently released an Open Images a dataset of ~9 million URLs to images that have been annotated with image-level labels and bounding boxes spanning thousands of classes.
However, when I was at the IBM workshop I asked a little Rasberry Pi TJbot “what do you see?”, it lit upmy electric crimson jacket in 5 different colors… Are we there yet to really use such tech for the design of any AI tech-driven solutions? For practical things based on such technology a lot of data, i.e. images or text, still needs to be fed in. In short, there is a need for the Custom Language model for using technologies like IBM Watson in the specific domain of the company or a start up.
There is a big market need for training the data to make a practical use of it.
Lastly, when English is a unifying language for us all, how do we bring such technologies to parts of the world where people speak other languages, let alone different dialects? How about colloquial language? Siri would not understand FD as Financial District, yet in the US we have the most advanced level of text and language recognition technologies, while other places are not there yet.
The majority of AI start ups are concentrated in the US, with a growing traction in Europe and Russia, but the rest of the world is still full of untapped opportunities.
All these unicorns, innovations, and ideas will be tailored and replicated everywhere. There will be Arab-market Instacart, and some entrepreneurs will take up a challenge to teach an autonomous car to drive in the luna-park-on-the-road looking Istanbul and Beirut. And teach chat bots Fusha or colloquial Arabic.
More human intelligence in training such data is needed. Human in the loop as called by CrowdFlower. Which is a job opportunity for many.
This is why more than ever I am inspired to work on the solution through my company TaQadam. We aim to bring remote work opportunities to the most vulnerable —the displaced, unemployed and excluded — refugee youth in the Middle East.
Cheers,
Karina
