Order from us for quality, customized work in due time of your choice.
Summary of Hurleys (2018) Journal Article
Artificial Intelligence (AI) has been termed as the science of teaching computers to accomplish tasks that can be completed through human intelligence. The Pentagon has viewed that incorporating AI is crucial in improving future military activities (Hurley, 2018). However, to ensure the secure incorporation of AI into military activities, the Department of Defense (DoD) must address various challenges.
These challenges include cultural silos, information insecurity, the DoDs size and complexity, technological diversity, and inadequate instruction. The DoD is discovering ways through which it can provide enhanced support to the military affairs and enhance the situational awareness to assist the fighters. The security body has been finding ways through which AI could provide defense and protection of national security knowledge. The AI has the potential to be applied in many government sectors, such as agriculture, transportation, cyber security, and weather. DoD has integrated AI into the Pentagons Third Offset Strategy to help stabilize and the counterinsurgency of military exercises. The AI-learning-enabled machines help military officers make timely decisions to combat security threats.
First Main Point
The costs associated with gathering, treating, and curating data for AI services are extremely expensive, thus making it hard for many organizations to consider it practical. I found the argument important, as it sums up some challenges associated with incorporating AI into various processes. The tremendous cost of AI incorporation in military operations is one challenge experienced by DoD (Hurley, 2018). The other concern that comes along with this high cost is that the AI models are only good on the data to which they are trained. The assessment and preparation of the data used in training are also difficult for its users.
Second Main Point
AI has been viewed with the potential of improving the DoD activity in addressing cybersecurity attacks. I found the point crucial as it sums up the benefit that AI incorporation can help the DoD. These cyber security tasks include defending against attacks, identification of susceptibilities, detecting attacks, and patching the vulnerabilities (Hurley, 2018). The ability of AI to reduce the vulnerabilities associated with cyberattacks will serve as an invaluable item to DoD. Using machine learning still faces challenges such as data security and human interferences. The machine learning algorithms have not been adherent to ethical and legal norms. Additionally, the machine learning models are prone to leakage of the training data, thus driving them not to be fully reliant on security missions.
Third Main Point
Stakeholders expect the AI progress in national security to be reflected in three sectors including military supremacy, economic supremacy, and finally economic dominance. I identified it as a crucial point that shows the vital areas that the AI incorporation will best fit within the military field. Cyber threats linked to data contamination and evasion attacks can happen at the training and interference stages, respectively (Hurley, 2018).
These threats use the advantage that the adaptive manner of machine learning algorithms can affect decision integrity by changing data and training input. In the data poisoning attacks, there is an injection of poisoned data in the specified information making the training model learn the wrong thing. Here, an attacker inputs information with the wrong labels, so that the data is wrongly encoded by the learning system. The hackers find the improperly labeled data from various unreliable sites. In the evasion attacks, the attackers aim to craft data that is wrongly classified by the learning model. The success of machine learning is correlated to the assessment of security-sensitive applications linked to the adversarial data.
Reference
Hurley, J. (2018). Enabling successful artificial intelligence implementation in the department of defense. Journal of Information Warfare, 17(2), 6582. Web.
Order from us for quality, customized work in due time of your choice.