The Future of AI

Drawing from subject matter experts and industry best practice, we have compiled key insights into the current and future states of AI that are consistent with Government policy, Gartner and McKinsey.​

This introspective was also based on a 2020 Keynote: Latest Advances in Artificial Intelligence from Richard Kimber at the Macquarie University Incubator, leveraging the proprietary daisee transcription and analysis solution.

​Fundamentally, AI will be a part of the future ways of working and companies must proactively approach the challenge to emerge successful.​

AI is the next global Arms Race

An AI strategy is key to navigating the complexity of the fourth industrial revolution.​

​While the assumption would be to consider the United States or China as leaders, it is evident the United Kingdom has positioned themselves to confront the emerging grand challenges of:​

​1. AI and data revolution;

​2.​ World leader in mobility; ​

​​3. A shift to clean growth;

​​4. Meeting the needs of an ageing society

Key Findings

1. From mobile to AI first: FAGMA (Facebook, Apple, Microsoft, Google, Amazon) companies have enjoyed financial success that has accelerated in the pandemic because they focused on developing AI strategy at all levels of the business

​ 2. Unstructured data: is the key challenge to overcome for machine learning to be truly effective in optimizing productivity ​

3. AI Service providers: Rather than investing intensive resources into internal AI, there will be emerging AI SAAS firms to affordably take advantage

- The FAGMA companies have all consistently chosen to be AI first in their strategy.

- The mystery behind ‘innovative and secret’ algorithms is now commonly understood.​

- The key point of differentiation is the approach to understanding and using data, and this is where FAGMA dominates.​

- Concerningly, the USA and China are the major players and there is a significant lack of Australian presence. ​

- The pandemic has accelerated the impact of an AI first approach, and the FAGMA performance shows in their financial success globally.​

What can the public and private sectors learn?​

- Become proactive, rather than reactive to encourage innovation.

​ - Move towards becoming a “Tech-Nation” that defines a technology strategy for the whole country.

​ - A committed partnership in innovation between businesses and universities to solve tomorrow’s challenges.​

AI vs Machine Learning

- AI and machine learning are not just a new kind of software but are completely new and unique.​

- Machine learning is a sub-set of AI that trains models to predict outcomes. AI machine learning recognises patterns and the more data that you serve the algorithm, the better is gets at pattern recognition.​

- This is very different to normal, traditional software that uses rules-based logic in a very systematic way to determine if this happens, do that.​

- Machine learning may be very accurate, up to 95%, but will never be perfect. However, humans are not perfect; it is important to calibrate AI to what a human can do, and in this measure, AI can replicate a human level of accuracy.​

What is the AI hype?

- AI has been over-hyped in the media repeatedly since 1956, but we are reaching a tipping point - as Gartner identifies.​

​- Popular visions of AI reaching Singularity is unrealistic, however, real prediction is possible and the next battleground for companies and countries.​

- AI has the capacity to augment people and their performance, as we can spend more time being productive.

​ - While the results of some elements such as chatbots are unsatisfactory, there is a new horizon emerging in computer vision.​

Why is AI Important

- Analytical tasks and finding patterns in data that was done by software, can be done more efficiently and at lower cost with AI​

- Advances in algorithms, training data, GPU’s and Cloud services along with greater interest and investment have created the perfect conditions to finally realise change.​

1. Innovation: New products and services will emerge.​

2. Efficacy & Velocity: Productivity and tasks will be achieved far more effectively and quicker.

3. Scalability: Tasks will be extended to more participants​

The opportunity of unstructured data

- Computer vision is the key to understanding abstract concepts that work with AI and Machine Learning.​

- In 2013, there was a breakthrough – we could convert words into representation or vectors (Word2Vec) that explore the common relationships in words.​

- This allows AI to understand the context and meaning behind the words in a numerical sense.​

- It changed the way AI can evaluate abstract concepts into close approximations of meaning.

This has opened extensive new possibilities with understanding the data. However, Most data is unstructured - just like our world. Language is the same, and for AI and machine learning to work, they need structure.​ ​

- 80% of corporate data will be unstructured by 2025 (IDC)

​- This is where Machine Learning and AI can make the biggest impact​ ​

- CRM systems will become autonomous in detecting information relevant to your customers rather than relying on static data​

- The data requires more storage, and it is harder to manage and protect with legacy solutions.​ - The amount of data is also growing, expected to reach 135 Zetabytes by 2025 (Data Age).​

- The magnitude of this would be to imagine attempting to collect and categorise every grain of sand on planet Earth for every Zetabyte of data!

Why AI is essential for the success of businesses​

- No longer relying on historic data but rather, AI can be used to predict future opportunities and threats​

- Optimisation – AI will not only have a dramatic impact on how businesses develop their go-to-market strategy but how to optimise it across different segments​

- Churn prediction using insights from AI and machine learning – it is far easier to retain a customer than earn a new one​

What can be done?​

- Necessary Skills ​ Awareness is key to putting AI first. While technical knowledge at every level isn’t needed, CEO and leadership need to join the AI conversation rather than delegating to R&D. ​​

- Identifying relevant use cases​ Knowing when, where and how to apply AI is the most important barrier to overcome.​

- Defining AI Strategy​

By understanding AI at a leadership level, companies can develop effective strategies to realise the rewards.​​

How will AI Services work?​

- Like Software as a Service products (SASS), AI will follow a similar approach.​

- AI will enable businesses to increase the velocity of innovation and continuously evolve through rapid release cycling.​

- Fundamentally, AI as a service lowers the cost of entry for customers because it removes the costly development component and can be purchased or rented.

- There are also several downstream integrations.​

We are daisee

We are daisee, a conversational middleware that automates the quality management process.

​Fundamentally we deal with unstructured interaction data and create structure in it to assist our clients with unearthing actionable insight.

​Recognised domestically and abroad, daisee is named as a Cool Vendor by Gartner, and in the KMPG Fintech Global 100, one of 7 Australian companies.

​Watch the full presentation here​