Updated: Nov 4, 2021
A recent BBC article asked, Are we on the cusp of an ‘AI winter’?
Describing the last decade as "a big one for artificial intelligence", the article goes on to claim that "researchers in the field believe that the industry is about to enter a new phase."
This new phase is predicted to be an ‘AI winter’ - an inflection point where AI research, investment and funding fall into a period of decline.
Likewise, peaks in AI interest and investment are known as ‘AI summers’.
AI WINTER 1
The term 'artificial intelligence' was coined in 1956 at the Dartmouth Summer Project, driven by the hypothesis:
“Every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it”.
This kickstarted the first wave of AI interest and investment from the 1950s to the early 1970s. This first AI summer marked interest and funding from government, academic, and military sources such as ONR, ARPA and the U.S. Navy.
Initial progress and enthusiasm in AI stalled, however, with disappointments in machine translation - the British Science Research Council’s bearish Lighthill report triggering the first AI winter around 1973, with AI investment and innovation falling into decline.
AI WINTER 2
By the 1980s, the pendulum began to swing the other way, with AI implemented in financial planning, medical diagnosis, geological exploration, and microelectronic circuit design.
In 1984, Business Week declared “AI: It’s Here”, explaining it is:
“...now possible to program human knowledge and experience into a computer. Artificial intelligence has finally come of age.”
Innovation in this second AI summer was characterised by ‘expert systems’, using the ‘top-down’ thinking of human experts to create ‘if-then’ rule sets. These expert systems leveraged desktop computers and cheaper servers to deliver results previously limited to expensive mainframes.
In effect, corporate cash drove the second AI summer in pursuit of business solutions, with rising expectations eventually hitting a brick wall in innovation. With 1980s expert systems highly dependent on data, storage proved too expensive to be practical. With no internet, no cloud and limited processing capability, companies proved unable to share data at scale.
As a result, corporate AI became siloed, eventually dampening further investment, with initial high hopes failing to be fulfilled. This led to the second AI winter, with DARPA concluding that AI research delivered:
“...very limited success in particular areas, followed immediately by failure to reach the broader goal at which these initial successes seem at first to hint...”
AI WINTER 3?
Many regard January 2007 as the beginning of the third AI summer, marked by Steve Jobs launching the iPhone.
Thirteen years on, is another AI winter coming?
We think the appetite's still there for risk, investment and innovation.
Japan's Society 5.0 play and Artificial Intelligence Technology Strategy announcement in 2017, supplemented in 2019 by a $1.4 billion injection into The Innovation Network Corp. of Japan, with eventual AI investment expected to be north of $4 billion.
In 2017, China pledging to dominate the AI landscape by 2030 with its own national strategy - with academic and commercial organizations in China working closely with the military on AI projects they call 'military-civil fusion'.
The British government announcing AI investment of more than $1.2 billion in 2018.
The French announcing $1.6 billion in AI funding in 2018
The European Commission expected to launch its AI regulatory guide in February 2020.
According to the Gartner CIO Agenda Survey, 48% of global CIOs will have deployed AI in 2020, with AI judged the most disruptive technology in IT budgets of over $67 billion.
Likewise, Gartner find that half of AI investments will be quantified and linked to specific key performance indicators to measure return on investment by 2024.
CLOUDS, MICRO APPS AND ZEROES
At the same time, however, relatively few organizations can currently claim a mature state of AI adoption.
In effect, the unique nature of every company is reflected in a unique AI maturity level, plus a unique approach to AI planning, management and execution.
At 22i, our strategy is to help organizations cut through the challenges of AI adoption, to simplify business process and efficiently roll out AI at speed and scale.
Strategically, this allows our clients to benefit from:
Lean functionality via relatively modest developer resources.
Quicker build, rollout and execution.
Modular, minimal and flexible solutions by design – meaning our AI can be tailored to specific tasks and processes.
Efficient deployment across various clouds, devices, platforms and business siloes.
Zero latency in process.
Zero distance to information.
Zero disruption to business operations.
REALITY OVER RHETORIC
At 22i, we make it possible to roll out AI without complexity - to plug the gap between AI rhetoric and reality.
We achieve this by stripping things back and focusing on the basics:
What's the key business issue here?
What's the critical business path?
Where are the pinch points?
What's the business objective?
What are the key performance indicators to measure success / return on investment?
This strategy allows us to deliver solutions that are easy to deploy, do not demand specialized AI skills from their users and ensure minimal disruption to the client's existing business.
At the same time, our AI transcends data siloes and integrates seamlessly with existing systems.
With a strategy more about the scalpel than the sledgehammer, our AI's designed to be future-proofed for the next bear market.
So, if winter is coming, let it come.
We think it will be a mild one.