Seven ways artificial intelligence could evolve
The world of artificial intelligence is only just beginning.
Given all the innovations taking place in the AI world, the possibilities that it creates are seemingly endless. Yet with increasingly prevalent concerns about ethics and the responsible use of AI, many business and information technology leaders are looking to better understand how this technology will affect organizations and society, both now and in the future.
Looking at future scenarios for how AI could evolve can help IT leaders demystify this emerging technology and better understand both its possibilities and its limitations. Here are seven possible future AI scenarios, ordered from those that may have an impact in the nearer term to those that will require many years before potentially becoming reality:
Augmented intelligence: complementing human strengths
Augmented intelligence is the idea of taking human intelligence as a starting point and supporting it through intuitive, yet non-invasive, user interfacing – such as holograms or interactive visualizations.
Examples in practice already exist, including augmented diagnosis for medical specialists or contextual recommendations for call center agents. Further development will require advances in human-machine interfacing, but this innovation will likely have significant business impacts in the near-term.
Generative AI: improving agency and emotion
Generative AI is the ability for machines to produce completely original and realistic structures, designs and even art. The potential of this technology is in already unfolding – being used, for example, to design new materials in engineering or augment development of new protein structures for vaccine production.
However, generative AI also comes with limitations and concerns given its potential for malevolent use, such as creating deepfakes or other fraudulent content. Regulatory or legislative hurdles could also hinder advances in this area, and developers will need to carefully consider AI ethics in the research and design phase.
Composite AI: going beyond machine learning
Composite AI is the combination of multiple AI techniques, such as machine learning and graph analytics. It is often used to combine data with other knowledge sources, such as human expertise and causal reasoning, enabling the development of a new generation of AI solutions.
For example, a company can use composite AI to build a more accurate predictive maintenance solution, which not only relies on sensor data but also experience-based heuristics or physical engineering models. This scenario is already unfolding, likely to have an impact within three to five years.
Steroid AI: more powerful hardware for more intelligence
The main idea behind steroid AI is improving the capacity of intelligent systems by using increasingly powerful hardware, such as neuromorphic hardware or quantum computing. Such advances will not change AI in any fundamental way; rather, it will allow companies to build AI solutions that are extremely fast and can encapsulate ever more data and knowledge.
For example, a virtual customer assistant may now have the computing capacity to respond to questions about product A and B, but with more powerful hardware, it could also cover products C, D and E. Advances in this area are ongoing, but the timeframe of its impact is uncertain. Neuromorphic computing is projected to have an impact within three to five years, while quantum and other more advanced systems will take longer.
Transcendent intelligence: synergy of humans and AI
Like augmented intelligence, transcendent intelligence takes human intelligence as a starting point, but in this case, it complements human intelligence using invasive human-brain implants.
While some early prototypes of this technology exist, such as artificial limbs or thought-controlled airplanes, these instances will require years of research to work effectively and responsibly. Significant progress is needed in neuroscience and psychology for deeper brain-machine integration, as well as in studying potential ethical and mental health implications.
Autonomous AI: leaping toward adaptability
In the future, AI could be used for truly adaptive physical or virtual robotics. This is not to be compared with the current generation of virtual assistants, drones or robots used in industries such as manufacturing. These future robots are more versatile, more multipurpose, more flexible and able to operate autonomously.
There are numerous advancements and ethical matters to consider, so autonomous AI is projected to be a longer-term development, requiring at least another decade before it becomes widely applicable.
General AI: an elusive pipe dream
The final scenario is the elusive concept of “general AI,” also known as AGI. This is the kind of AI that is equal to or better than human intelligence – what’s seen in movies such as “Star Trek” or “The Terminator.”
Most AI scientists today agree that AGI is still many decades away and will require fundamental breakthroughs in neuroscience and computer science, as well as massive advances in compute power. It will also require the mitigation of profound societal and human implications that are yet to be discovered.
As AI research and development progresses, IT leaders should prepare for new innovations by starting to build new practices and skills that go beyond data-driven machine learning. Look to apply new forms of AI in more sophisticated business value cases, such as augmented product research and development, personalized customer interaction and business process automation. Frequently revisit and adapt your AI strategy to reap the benefits of new AI advances and to identify newly required skills and capabilities.
Not least, manage the growing impact of AI by developing and deploying ethical awareness, regulations, policies, governance and practical guidelines first – not after AI solutions are deployed.
Dr. Pieter J. den Hamer is a senior research director at Gartner, Inc., covering artificial intelligence and related topics such as data science, optimization and decision intelligence. He wrote this article for SiliconANGLE. He and other Gartner analysts will present on these topics during Gartner IT Symposium/Xpo 2021, taking place virtually in the Americas Oct 18-21.
A message from John Furrier, co-founder of SiliconANGLE:
Show your support for our mission by joining our Cube Club and Cube Event Community of experts. Join the community that includes Amazon Web Services and Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger and many more luminaries and experts.
We really want to hear from you, and we’re looking forward to seeing you at the event and in theCUBE Club.