Latest Trends in Artificial Intelligence

Verified

Added on  2023/04/24

|6
|1503
|424
AI Summary
This report discusses the latest trends in Artificial Intelligence (AI) including reduced time consumed in training, AI transparency and responsibility, automation of tasks, specialized hardware for sensing and model interference, hybrid models, and investments in new techniques and tools.

Contribute Materials

Your contribution can guide someone’s learning journey. Share your documents today.
Document Page
Running head: LATEST TRENDS IN ARTIFICIAL INTELLIGENCE
LATEST TRENDS IN ARTIFICIAL INTELLIGENCE
Name of Student
Name of University
Author’s Note

Secure Best Marks with AI Grader

Need help grading? Try our AI Grader for instant feedback on your assignments.
Document Page
1LATEST TRENDS IN ARTIFICIAL INTELLIGENCE
Introduction
Every media has been highlighting news regarding the dangers imposed by machine
learning, fake new about these have been increased a lot since few years. Yet the technologies
have continued to develop in automation as well as machine deception, they would help in
continuing to structure the usage of Artificial Intelligence (AI) over the upcoming years (Fast &
Horvitz, 2017). According to a recent study it has been found that the companies who have
invested in AI have obtained a financial return on the investments made by them. This report
discusses regarding the latest trends of AI in details in the discussion part.
Discussion
AI includes various inspiring as well as significant trends, these trends are as follows
Reduced time consumed in training: academic tasks that are performed using AI
usually aim in reducing the time as well as the computing power consumed for training a model
in an effective manner. With this the goal is to make the technology usable as well as affordable
in day-to-day work (Lorica, 2018). The technology regarding artificial neural networks has been
known for a long time, this works if there are sufficient tools for computing machine learning
models. A specific way by which it can be ensure that these cores are present include designing
of powerful hardware, the limitations of this must be considered as well. One more approach
includes designing new models as well as improve the existing ones for making it less
computing. The time required to train a model can also be reduced by optimizing the hardware
infrastructure that is required (Budek, 2019). Google Cloud Platform offers a cloud based
environment that helps in building machine learning models, it does not even require the
investigation regarding on-prem infrastructure.
Document Page
2LATEST TRENDS IN ARTIFICIAL INTELLIGENCE
Diagram 1: reduction of time required for training
(Source: 10 trends of Artificial Intelligence (AI) in 2019)
AI transparency and responsibility: machine learning has an ever-growing impact on
the growth of the business along with social as well as legal impacts. The main question that
rises with implementation of AI is that who must be held responsible for the failure of the
technology. The main issues here includes the hidden bias in various data sets, this creates a
problem for every company that uses AI in order to power up their regular operations (Jordan &
Mitchell, 2015). An example is Amazon that had used AI for the purpose of preprocessing the
resumes. Women who applied for the job were considered before the resumes that had an
experience of around 10 years. As organizations adopt machine learning the transparency of AI
also increase. According to various studies, the mistakes performed by the technology depend on
the culture in which in the responder has grown up (Zhiming, Xiaxia & Xuan, 2017). In extreme
conditions the driver is responsible for the choices.
Document Page
3LATEST TRENDS IN ARTIFICIAL INTELLIGENCE
Automation of tasks: automation takes place in various stages; the complete process of
automation might be a difficult process because it consists of many tasks and workflows those
results in partial automation. According to McKinsey less than 5% of activities can be automated
with the use of present technology (Falomir, Gibert & Plaza, 2018). Whereas around 60% of
activities can have almost 30% of their constituent tasks automated. Instead of waiting for
complete automation, the competition would be responsible for driving organizations for
implementing partial automation solutions. The success of these partial automations would help
in future developments.
Specialized hardware for sensing and model interference: the resurgence in the
concept of deep learning had started in the year 2011 with various record setting models in the
computer vision and speech. There is sufficient scale for justifying specialized hardware, for
example Facebook makes a huge number of predictions every day (Kaenampornpan, Malaka &
Nguyen, 2018). Google also has sufficient scale for justifying the production of specialized
hardware. It uses tensor processing units on their cloud since the year 2018. Various companies
as well as startups in US as well as China are working on hardware which helps in targeting
model buildings as well as interference, both in on edge devices as well as data centers.
Hybrid models will stay important: as deep learning continues driving the research
procedure, most of the solutions are hybrid in nature. In 2019, the organizations would hear more
regarding the important roles of various other components as well as methods, these methods
also include the ones that are model-based such as Bayesian inference, evolution, tree search,
simulation platforms, knowledge graphs and many more (Lorica, 2018). Many new
developments might also be seen in the topic of machine learning methods that are not actually
based on various neural networks.

Secure Best Marks with AI Grader

Need help grading? Try our AI Grader for instant feedback on your assignments.
Document Page
4LATEST TRENDS IN ARTIFICIAL INTELLIGENCE
Investments in new techniques and tools: the tools that are utilized for machine
learning development would need to be considered for the effectiveness of data, model search
and experimentation, monitoring and model deployment (Budek, 2019). Companies usually
begin to look in the tools for lineage of data, analysis and metadata management, effectively
utilizing the computing resources, effective model searching along with hyper parameter tuning.
In the year 2019, more tools can be expected to ease the procedure of development as well as
actual development of machine learning and Artificial intelligence to various services and
products.
Conclusion
The field of Artificial Intelligence has numerous latest trends that had been detected.
Some trends include new technologies would enable partial automation of numerous tasks,
artificial intelligence in enterprises would build upon the existing applications that re analytical
in nature, UI/UX design would be critical, hardware would be more specialized for the purpose
of sensing, model interference and model training, hybrid models would stay important,
investments would be done in new processes as well as tolls and techniques, challenges in
machine deception would increase, questions would be raised around safety as well as reliability
and organizations would be able to access more data which would help them to take the
advantage of data that they did not actually generate. Some more trends include virtual assistants
and chat bots ride the lightening, reduction in time consumed in training, autonomous vehicles’
rising speed, artificial intelligence and machine learning would be democratized as well as
product ionized, AI responsibility as well as transparency and many more. This report discusses
regarding some of these trends in details.
Document Page
5LATEST TRENDS IN ARTIFICIAL INTELLIGENCE
References
10 trends of Artificial Intelligence (AI) in 2019. (2019). Retrieved from:
https://becominghuman.ai/10-trends-of-artificial-intelligence-ai-in-2019-65d8a373b6e6
Budek. K, (January 9, 2019). Five top artificial intelligence (AI) trends for 2019. Retrieved from:
https://deepsense.ai/ai-trends-2019/
Falomir, Z., Gibert, K., & Plaza, E. (Eds.). (2018). Artificial Intelligence Research and
Development: Current Challenges, New Trends and Applications (Vol. 308). IOS Press.
Fast, E., & Horvitz, E. (2017, February). Long-term trends in the public perception of artificial
intelligence. In Thirty-First AAAI Conference on Artificial Intelligence.
Jordan, M. I., & Mitchell, T. M. (2015). Machine learning: Trends, perspectives, and
prospects. Science, 349(6245), 255-260.
Kaenampornpan, M., Malaka, R., Nguyen, D. D., & Schwind, N. (2018). Multi-disciplinary
Trends in Artificial Intelligence. Springer International Publishing.
Lorica. B, (November 22, 2018). Artificial intelligence trends for 2019. Retrieved from:
https://www.itproportal.com/features/artificial-intelligence-trends-for-2019
Zhiming, Y., Xiaxia, T., Xuan, Q., Fei, Z., & Yuanmei, D. (2017). The Connotations, Key
Technologies and Application Trends of Educational Artificial Intelligence (EAI):
Interpretation and Analysis of the Two Reports Entitled “Preparing for the Future of
Artificial Intelligence” and “The National Artificial Intelligence. Journal of Distance
Education, (1), 3.
1 out of 6
circle_padding
hide_on_mobile
zoom_out_icon
[object Object]

Your All-in-One AI-Powered Toolkit for Academic Success.

Available 24*7 on WhatsApp / Email

[object Object]