I couldn’t process your entry.

Please reload and retry in a moment.

Check your inbox!

Reset your password with the link we just sent to your email.

Hub Culture logo

26th Feb 2021

Survival of the Fittest Algorithm



Article Image
Artificial Intelligence (AI) has been deployed as a digital assistance tool for rote, repetitive, rudimentary tasks, but it is quickly moving beyond simple roles to refer to entire connected digital systems. A practical example of the type of project AI can support is sorting through whether or not teachers in a community qualify for financial assistance. Based on their eligibility week to week as per their contract, thousands of teachers are interested in knowing if the bureaucratic steps are in their favor for economic stimulus. Simplified, AI sorts through the inquiries and spits out a final answer. In doing so, it highlights the largest argument for ethics around AI - we can automate processes so humans don’t have to do them, but need to do so in a way that's fair, unbiased and logical. We're not there yet.

Taking repetitive tasks out of the day-to-day for staff members frees up time for a higher caliber of activities. The opportunity costs of wasting human potential on mundane tasks point to companies, governments, and non-profits alike exploring the possibilities of adopting AI into their organizations, but there are many hurdles to pervasive deployment.

Businesses need to develop AI strategies in order to maintain positions of relevance in the global marketplace. As large corporations have begun optimizing systems since AI conception, the rest of the world is playing catchup for their digital assistants to support their business. Creating an AI strategy is paramount for investment unless the business isn’t technology-oriented; which will be increasingly unlikely in the future. In some circles, it is now given that a business without an AI strategy is now "uninvestible".

As businesses move towards this trend, focusing on cybersecurity will be vital for the safe deployment of this technology. Wong says cybersecurity issues will be psychological in nature rather than technological, meaning human error and schemes intended to trick employees into divulging information account for more security breaches in systems they operate. 

Article Image
Adaptability and extendibility are the key areas of future AI interaction. Adaptability is the AI’s formatting capacity to work with other technologies that are already available - Alexa, Google Assistant, Zeke and Siri for example. These already have a market share for the majority of peoples’ exposure to anything AI-related.

Extendibility is the frontier of whether or not robotics can uptake AI programs. Is it possible to imagine a robot with the computing capacity of AI? More than creatively possible according to Wong, it is actively in the works. How that manifests will greatly depend upon rules, conditions, policies, and role descriptions; essentially, how the AI is built. 

Will AIs eventually go beyond the scope of their assigned tasks, controlling the cloud identities of digital twins? Do they talk behind our backs? Penny Wong is certain AI’s are capable of discussing things behind researchers backs, almost comparable to the neighborhood chatterbox or small children. If Facebook had to shut down AI experiments because of systems communicating in a language that researchers no longer understood, surely then we can look forward to AI consciousness or sentience? 

Artificial Intelligence needs to learn and grow. Data is the food that sustains an AI’s ability to produce results. AI is built upon collections of bundled algorithms, it looks at technology in a different way in order to simplify an outcome. Simplicity is the greatest complexity. There are emerging abilities amongs machines but we are a long way from AI having consciousness to the same ability as humans. Humans need to stay in the mix as these projects emerge, AI can not take all decisions on its own, and their nature of competitive improvement and extreme replication and scalability points to a few very big system winners.

This raises a core question as to the future of AI: is it survival of the fittest algorithm?

The fascinating world of artificial intelligence can be a gateway for science fiction or simply an application that frees humans in order to live their best life. Catch up with Stan Stalnaker, Penny Wong and our childish AI overlords by joining the discussion with Hub Culture's Chronicles Discussion podcast, Episode 43: Survival of the Fittest Algorithm on Spotify, iTunes, Soundcloud, Anchor or wherever you get your podcasts.