Disclaimer: This is an example of a student written essay.
Click here for sample essays written by our professional writers.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UKEssays.com.

Proposal for Research on Ethics of Artificial Intelligence in Machine Learning

Paper Type: Free Essay Subject: Technology
Wordcount: 6424 words Published: 8th Feb 2020

Reference this

Abstract

Artificial Intelligence (AI) is a powerful science which utilizes sufficient methodologies, strategies, and systems to take care of unsolvable real world issues. Due to its relentless ascent towards the future, there are additionally a few discussions about its ethics and safety. Machine ethics is concerned about the results of conduct of machines towards human users and other machines. The significance of machine learning ethical contemplations will just develop as our dependence on technology increases. Particularly in conditions with legal and ethical outcomes, we should think about the after effects of these machines to guarantee decency. For each present day individual whose life includes the utilization of a gadget to get to the Internet, machine learning algorithms have turned into the texture of their experience. To help with deciding the best arrangement, we swing to machine ethics.

Key Words

artificial Intelligence, machine ethics, algorithms, ethical artificial intelligence, machine learning

Contents

1. Background…………………………………………3

2. Research questions……………………………………3

3. Literature Review……………………………………..4

3.1. Overview of Artificial Intelligence …………………………4

3.2. Current State of AI…………………………………..4

3.3. Ethics in Machine Learning and its Algorithm………………….5

3.3. Technical challenges in Machine Learning……………………6

4. Methodology Review…………………………………..6

4.1. Interviews………………………………………..7

4.2. Focus Groups……………………………………..7

4.3. Questionnaires……………………………………..8

4.4. Surveys………………………………………….8

4.5. Case Studies………………………………………9

4.6. Methods taken to achieve the Research Process………………10

4.6.1. Interview……………………………………..10

4.6.2. Survey……………………………………….10

5. Conclusion………………………………………….11

6. Project Plan…………………………………………11

7. Ethical Issues………………………………………..11

8. Consent Form………………………………………..12

9. Pilot interview questions………………………………..12

10. Pilot survey questions and design………………………….12

11. References………………………………………….13

12. Appendices………………………………………….15

 12.1. Appendix A………………………………………..15

 12.2. Appendix B…………………………………………16

12.3. Appendix C………………………………………..21

12.4. Appendix D………………………………………..23

12.5. Appendix E……………………………………….24

1.   Background

The emerging area of artificial Intelligence (AI) technology has grown unexpectedly in latest years, reinforced by way of big medical and technological progress in lots of fields: engineering, science, training, medicine, enterprise, accounting, finance, advertising, economics, inventory marketplace and regulation, amongst others (Halal,2003).The destiny relies upon on roles of many progressive fields in shaping the humankind’s existence. Artificial Intelligence (AI) is known as this kind of fields as making the sensible life more self-maintaining. This example is because of fast changes and traits in mainly computer and electronics oriented technologies. In time, there had been many tremendous upgrades making revolutionary jumps in these technology and besides from a few particular risks, they have got taken increasingly role inside the modern life, manner too many one in every of a kind benefits – blessings.

Get Help With Your Essay

If you need assistance with writing your essay, our professional essay writing service is here to help!

Essay Writing Service

With active roles of such technologies, human beings now have a higher existence and appears hopeful to the destiny. However it is continuously an active question these days that what if the destiny might not be in line with the humankind’s desire. Such anxieties are typically related to the lifestyles of the humankind and distinctive residing creatures over the world. As impact of the AI may be measured substantially from both micro and macro-views actively (Pavaloiu, 2016), it’s also a first rate research way to consider how can AI be utilized in a moral manner and saved efficiently. AI is now the reality, which need to be evaluated in a vital way for a higher destiny. Due to that, moral AI has end up a popular research studies in twenty first Century.

2.   Research questions

  1. What are all the ethics applicable for the machine learning using AI?
  2. How could we properly manage any ethical and legal risk arise from while using it?

3.   Literature Review

3.1.           Overview of Artificial Intelligence

Artificial Intelligence is the field of have a look at that seeks to explain and emulate intelligent behaviour in phrases of computational processes (Schalkoff, 1990). Its miles the era and engineering of making practical machines, in particular sensible laptop applications. It’s far some distance related to the identical mission of the usage of computer systems to understand human intelligence, however AI does not want to confine itself to techniques which are biologically observable. The core troubles of Artificial intelligence consist of programming computer systems for positive trends along with: knowledge, reasoning, trouble solving, perception, learning and making plans.

The beginning of artificial Intelligence in the year (1952-1956). It is normally conventional as having begun with the invention of robots. present AI and robotics components can provide physical structures, notion, motor control, navigation, mapping, tactical selection-making and long-time period planning (Berkeley, 2015).John McCarthy coined the term “artificial intelligence” (AI) in 1955, defining it as “the technology and engineering of making sensible machines”.

Even as AI is becoming even extra superior and integrated into our lives, worries approximately the ethical troubles in developing such era and the chance related to adopting it end up more essential (Taylor et al, 2016). In this proposal identify the principle moral concerns and ability dangers related to AI and propose methods to control and mitigate capacity existential dangers and maximize the nicely-being of the human race.

3.2.           Current State of AI

In latest years, there were maximum critical advances in developing the research and era of AI into performant merchandise with a focal point on numerous era. A machine will use such models to form predictions concerning future information, and take selections that are rational given these predictions (Zoubin Ghahramani, 2015). Uncertainty plays an elementary half altogether of this. Discovered information are often in line with several models, and thus that model is acceptable, given the information, is unsure.

Many aspects of learning and intelligence crucially rely on the careful probabilistic illustration of uncertainty. Probabilistic approaches have solely recently become a thought approach to artificial intelligence, robotics and machine learning (Zoubin Ghahramani, 2015). Even now, there’s arguing in these fields concerning however necessary it’s to completely represent uncertainty. For instance, advances exploitation deep neural networks to resolve difficult pattern-recognition issues like speech recognition, image classification, and prediction of words in text, don’t overtly represent the uncertainty within the structure or parameters of these neural networks.

3.3.           Ethics in Machine Learning and its Algorithm

Imagine, within the close to future, a bank using a machine learning algorithmic program to suggest mortgage applications for approval. A rejected human brings a cause against the bank, alleging that the algorithmic program is discriminating racially against mortgage candidates. The bank replies that this is not possible, since the algorithmic program is deliberately unsighted to the race of the candidates. Indeed, that was a part of the bank’s explanation for implementing the system. Even so, statistics show that the bank’s approval rate for black candidates has been steady dropping.

Artificial Intelligence Algorithms also play an important role in the modern society. Some challenges of machine ethics are very similar to several alternative challenges concerned in designing machines (Eliezer, 2014). Planning a robot arm to avoid crushing stray humans isn’t any a lot of morally fraught than planning a non-combustible couch. It involves new programming challenges, however no new ethical challenges.

Responsibility, transparency, auditability, honesties, certainty, and a bent to not create innocent victims scream with helpless frustration (Nick Bostrom, 2014). The entire criteria that apply to human’s activity social functions. Those criteria that has got to be thought-about in an algorithm supposed to exchange human judgment of social functions (Eliezer, 2014).

3.4.           Technical challenges in Machine Learning

Machine Learning is a lively field of look into, with a scope of territories for further improvement crosswise over various techniques and application zones. Current advances in the field have opened up further energizing exploration challenges. Most of the challenge in the form of technical based. For Examples, rising hardware to support machine learning systems and coming up with algorithms that don’t need large amounts of labelled data.

In extra to those strictly technical challenges, the addition of applications of machine learning has detached a brand new wave of analysis challenges, that relate to each technical advances within the field and social challenges associated with machine learning. A neural network is one of the approach to machine learning in this very tiny computational units are merged in a way that is enlivened by associations in the cerebrum.

The broad variety of day to day machine learning gives a benefits to the social and economic purposes. To achieve the benefits of machine learning it will depend on increasing the absorptive capacity of UK industry, so that it can utilize the machine learning. The ongoing achievement of machine learning in coordinating or notwithstanding enhancing human conduct for certain assignments, including picture and speech recognition, is expected in no little part to the ongoing blast of preparing information in these regions. The immense number of models on which to prepare the algorithms has been a basic piece of their improved executions.

4.   Methodology Review

There are different methodologies that can be utilized inside research process to effectively address an exploration question. Focus groups and interviews have become a popular strategy for gathering qualitative information. Focus groups are right now utilized as both an independent strategy and in blend with studies and other research techniques, most quite individual, in-depth interviews. Each and every methodologies has an advantages and disadvantage. So it is very important to choose the research methods depending on the information gathered.

4.1.           Interviews

In social research there are numerous kinds of interviews. The most basic of these are unstructured, semi-structured and structured interviews (Catherine Dawson, 2002). In the unstructured kind of interview, the research endeavours to accomplish an all-encompassing comprehension of the interviewer’s perspective or circumstance. Analysts need to remain alert, perceiving significant data and testing for more detail. They have to realize how to carefully guide somebody once again from absolutely unimportant deviations. Moreover, it is critical to understand that unstructured interview can create a lot of information which can be troublesome to analyse. Semi-structured interview is mostly used in qualitative social research. This kind of interview is taken for the research to confirm continuity. Structured interview are mainly used for the market research purpose. It is based on the quantitative research and it may be conducted face to face or telephone interview (Dawson, 2002).

In the Face to face interview the main advantage of social cues. It can be like voice, body language, intonation etc. (Raymond Opdenakker, 2006). If the interviewer wants to know the character of the interviewee social cues are must. Suppose if the interview taken from the expert person the social cues is not much important. The main advantage of the interview is that the appropriate response of the interviewee is increasingly unconstrained, without an all-inclusive reflection. The drawback of face to face interview is time consuming. For the telephone interview the advantage is extended access to participants compare to the previous interview type. The disadvantage for the telephone interview is less chance for standardisation of the situation (Raymond Opdenakker, 2006).

4.2.           Focus Groups

Focus groups might be called discussion groups or group interviews. Various individuals are approached to come up in an interview to talk about a specific issues. The focus group method is now considered to be a very innovative research method (Ivana Acocella, 2012). It is one of the most suitable technique to achieve the cognitive goals of the research. It also gathering qualitative data, both among sociologists and over a wide scope of scholarly and connected research areas (Morgan, 1996). Focus groups may be video or tap recorded (Dawson, 2012).

The advantage of focus groups can be amplified through cautious consideration regarding research configuration issues at both the task and the gathering level. Significant future headings include: the advancement of measures for announcing focus group research, progressively methodological research on focus groups (Morgan, 1996), more regard for information examination issues, and greater commitment with the worries of the exploration members. The disadvantage of focus groups can be some participants may be not comfortable in a group setting and nervous about speaking in front of many participants. The main drawback is very tough to extract the individual views in the analysis (Dawson, 2012).

4.3.           Questionnaires

A questionnaire is a research instrument comprising of a progression of inquiries to accumulate data from respondents. Questionnaires can be delegated both, quantitative and qualitative strategy relying upon the idea of questions. In particular, answers got through closed-ended questions with different decision answer choices are analysed utilizing quantitative methods and they may include pie-charts, bar-graphs and rates (Robinson, 2018). Answers acquired to open-ended questions are broke down utilizing qualitative methods and they include dialogs and critical analyses without utilization of numbers and computations.

Find Out How UKEssays.com Can Help You!

Our academic experts are ready and waiting to assist with any writing project you may have. From simple essay plans, through to full dissertations, you can guarantee we have a service perfectly matched to your needs.

View our services

Questionnaires can be classified into three forms such as open-ended, closed-ended and combination of both (Dawson, 2002). Advantage of questionnaires incorporate expanded speed of information accumulation, low or no cost necessities, and more elevated amounts of objectivity contrasted with numerous elective techniques for essential information gathering. In any case, questionnaires have certain disadvantage, for example, determination of arbitrary answer decisions by respondents without legitimately perusing the questions (Robinson, 2018). Besides, there is typically no plausibility for respondents to express their extra contemplations about the issue because of the nonattendance of an applicable inquiry.

4.4.           Surveys

Survey research is additionally characterized as a technique for graphic research utilized for gathering essential information dependent on verbal or composed correspondence with an agent test of individual or respondents from the target population. According to Kerlinger (1993), characterized the techniques for information accumulation in survey research as kinds of surveys like individual interview, Mail survey, Panel, phone and perceptions.

The advantage of the survey is high representativeness it can able to extract data that are near to the exact attributes of the large peoples (Kelley et al, 2003). The important benefit is low cost, convenient data gathering, Good statistical significance and precise results. The drawbacks in the survey has difficult to understand the large number of varied opinion from the user and huge amount of fake information that leads to affect the overall reasons for which a survey is created(Kelley et al, 2003).

4.5.           Case Studies

Case study inquires about, through reports of past studies, permits the investigation and comprehension of complex issues. It tends to be viewed as a vigorous research technique especially when an all-encompassing, in-depth investigation is required. Through case study techniques, a scientist can go past the quantitative measurable outcomes and comprehend the conduct conditions through the entertainer’s viewpoint. By including both quantitative and qualitative information, case study clarifies both the procedure what’s more, result of a wonder through complete perception, remaking and examination of the cases under scrutiny (Tellis, 1997).

According to Zainal (2007), the case study can be classified into three types such as exploratory, descriptive and explanatory case studies. Exploratory case study on individual’s reading process can be ask general questions. Descriptive case studies is used to depict the natural phenomena which happen inside the information being referred to (Grassel & Schirmer, 2006), for example, what various procedures are utilized by a reader and how the reader use them.

The advantage of case study is within the circumstance in which the movement happens and it also allows both quantitative and qualitative analyses of the information. The definite qualitative records frequently delivered in case studies not just investigate or then again depict the information, in actuality, condition, yet additionally help to clarify the complexities of real-life circumstances which may not be caught through trial or survey research The disadvantage of the case study is difficult to conduct and producing a large amount of documentation(Zainal, 2007).The main drawback in the case studies is its reliance on a solitary case investigation making it hard to achieve a generalising conclusion (Tellis, 1997).

4.6.           Methods taken to achieve the research Process

4.6.1.  Interviews

Interviews can be characterized as a qualitative research method which includes directing serious individual interviews with few respondents to investigate their points of view on a specific thought, program or circumstance. One reason as why this methodology was picked is on the grounds that it very well may be utilized to assemble open-finished replies by which workers not replied in a survey. It will likewise give the members (workers) the opportunity to develop their responses to the inquiries being posed enabling them to answer them in brief, and then again enable them to pose any questions identified with the topic. Likewise the member will be interview in their characteristic workplace and via completing a face to face interview will enable the specialist to watch the responses of the members when they are responding to the questions.

In the interview the questions like “How do we protect access to AI from bad conductors”? this type of questions enable the researcher to get as much data as they can to respond to the research question, and furthermore empower the member to express their worries in detail. Besides, if the member does not comprehend the inquiry they are being posed, they can demand the questioner (analyst) to disclose the inquiry to them with the goal that they completely comprehend the inquiry which will imply that reliable can be accumulated. The interviews taken from various geographical areas and universities around twenty members. The interviews will be taken approximately twenty to twenty five minutes for over fifteen days.

4.6.2.  Surveys

Survey can be clarified as addressing people on a theme or points and after that depicting their responses. In business ponders survey for essential information gathering is utilized so as to test ideas, reflect mentality of individuals, set up the dimension of consumer loyalty, lead division explore and a lot of different purposes. Survey technique can be utilized in both, quantitative and qualitative investigations.

Survey methods seeks for two primary purposes: Portraying certain angles or attributes of populace and testing speculations about nature of connections inside a populace. Surveys can be directed quicker and less expensive contrasted with different techniques for essential information gathering such as perception and analyses. Essential information assembled through survey are generally simple to break down. Two hundred participants from three various universities from different geographical area will be asked take an interest in finishing the review.

5.   Conclusion

The emerging technology of Artificial Intelligence has risk and ethical concerns if it is managed properly it will lead to better quality of life for human society. Machine ethics confounded and multifaceted issue. Be that as it may, in the event that we hit the nail on the head, we will release the full advantage of AI for mankind. The proposal demonstrates that we ought to not really believe a machine learning algorithm. It is obviously not moral to utilize a machine to make decision when we don’t confide in it for using good decision. Various procedures will be utilized to respond to the research question and the venture will be separated into errands so as to finish it effectively.

  1. Project Plan

The target of a project plan is to characterize the way to deal with be utilized by the project group to convey the planned undertaking the executives extent of the project. Refer to Appendix A in the Appendices area to see the Gantt outline; here you will see the research project is separated in tasks with proper dates set for the timescale and culmination of each undertaking.

  1. Ethical Issues

With the goal for research to result in advantage and limit danger of mischief, it must be led ethically. By getting ethical endorsement you are exhibiting that you have clung to the acknowledged ethical principles of a certified research contemplate which could build enlistment potential. In the Appendix B ethical review form is provided to expressing the conceivable ethical issues identified with this project.

  1. Consent Form

As information about individuals will be accumulated and used to respond to the research question. In this way, every member participation will be given a consent form which need to sign so as to participate in the research process. Kindly allude to Appendix C in the Appendices area to see a copy of the form.

  1. Pilot Interview questions

A pilot interview is planned not for information accumulation in that capacity but rather as a guide to the structure of later research. Early pilot interview will generally be unstructured, language and ideas, however later ones may test the reasonableness of institutionalized instruments. An Interview will be one of the ways to deal with accumulate information so as to respond to the research question. So, a pilot talk with plan with inquiries questions is given in the Appendix D.

  1.         Pilot Survey questions and Design

Survey may feature areas where certain inquiries don’t restore the data that you need. A Survey will be another methodology used to accumulate information so as to address the research question. Hence, a pilot survey with questions is given in the Appendix E to refer.

11.         References

  • Anderson et al, (2007) – The status of machine ethics: a report from the AAAI conference. Minds and Machines, 17(1): 1-10.
  • Alice Pavaloiu, and Kose, U (2017) – Ethical Artificial Intelligence an Open Question. Journal of Multidisciplinary Developments. 2(2), 15-27, 2017.
  • Armstrong, S., Bostrom, N. and Shulman, C. (2015) – A model of scientific intelligence development. AI & SOCIETY, 31(2), pp.201-206.
  • Bryson J, and Alan Winfield (2017) – Standardizing Ethical Design for Artificial Intelligence and Autonomous Systems. Published by the IEEE computer society.
  • Bottou, L. (2013) – From machine learning to machine reasoning. Machine Learning, 94(2), pp.133-149.
  • Davies, J. (2016) -Program good ethics into artificial intelligence. Nature.
  • Davis, E. (2015) – Ethical guidelines for a super intelligence. Artificial Intelligence, 220: 121-124.
  • Eaden, J., Mayberry, M. and Mayberry, J. (1999). Research strategies: Questionnaires: the employment and abuse of social survey methods in medical research. Postgraduate Medical Journal, 75(885), pp.397-400.
  • Fuchs, L. and Fuchs, D. (2017). Principles for research-Based Practice in the Schools: A Case Study. Focus on Exceptional Children, 33(6).
  • Giovanni Acampora, Diane J. Cook, F, and Parisa Rashidi (2013) – A Survey on Ambient Intelligence in Healthcare. Proceedings of the IEEE | Vol. 101, No. 12, December 2013.
  • Horel, E. and Giesecke, K. (2018) -Towards Explainable AI: Significance Tests for Neural Networks. SSRN Electronic Journal.
  • Nath, R. and Sahu, V. (2017) -The downside of machine ethics in computing. AI & SOCIETY.
  • Nick Bostrom and Eliiezer Yudkowsky (2011) – The artificiali intelligencei ethics in future. (Cambridge University Press).
  • Prescott, T. and Szollosy, M. (2017) -Ethical principles of robotics. Connection Science, 29(2), pp.119-123.
  • Rina Fitriana, Eriyatno, and Taufik Djatna (2015) – Progress in Business Intelligence System research: A literature Review. International Journal of Basic & Applied Sciences IJBAS-IJENS Volume: 11 No: 03.
  • Royakkers, L. and van EST, R. (2015)-A Literature Review on New Robotics: Automation from Like to War. International Journal of Social Robotics, 7(5), pp.549-570.
  • S. A. Oke (2008) – A Literature Review on Artificial Intelligence. International Journal of Information and Management Sciences Volume 19, Number 4, pp. 535-570.
  • Taylor Meek et al, (2016) – Managing the Ethical and Risk Implications of Rapid Advances in Artificial Intelligence. Technology management of social innovation (PICMET’16).
  • Theodorou, N. (2013). Qualitative Research Methodologies, Processes Qualitative Research in the Health Sciences – Methodologies, Methods and Processes. Nursing Standard, 28(3), pp.28-28.
  • Waser, M. (2015) – Designing and Enforcing a Coherent System of Laws, Ethics and Morals for Intelligent Machines (Including Humans). Procedia Computer Science, 71: 106-111.

APPENDICES

12.         Appendices

12.1.      Appendix A

 A Gantt chart that provides a comprehensive time line for the work

 Project Plan – Gantt chart

12.2.      Appendix B

 

 

12.3.      APPENDIX C

The Ethics of Artificial Intelligence in Machine Learning Research Consent Form

Dear Participant,

The reason for this interview and survey is for an exploration reason just as it is a piece of my research project. “The ethics of Artificial Intelligence in machine learning”. The point of this exploration project is to examine the concern representatives have with respect to the information that is gathered about them through artificial intelligence in machine ethics, and whether they know about the artificial intelligence usage in the real world whether it is good or bad in the future.

I am welcoming you to participate in an interview and survey in which you will be asked to answer a progression of inquiries; you may wish to finish them both around the same time or you may to finish the review time permitting and send it back to me. The meeting research session will include just two individuals, me and the member (yourself) and it won’t be recorded in any sort of way, just your answers will be brought down on paper by me, and the interview will last around 25 minutes.

Your take part in this process is wilful and you reserve the option to pull back whenever. You will get the opportunity to keep a signed copy of consent form in the event that you wish to partake in this research, and any data recorded from this exploration won’t be passed on to any outsider. Every one of the information and data gathered from this research is completely safe and may be put away on my own electronic devices.

The individual information accumulation from this research will be limited and the individual information which is fundamentally required will be requested generally not. Your information might be utilized with this project yet won’t have any of your own subtleties included.

When the research project is finished, the information and data accumulated about the member will be kept for a greatest time of 1 year yet, close to that, and the whole information and data gathered will be completely annihilated.

By signing this consent form you determine that you understand the following:

I keep in mind that even though I agree to take part now, I can withdraw at any time or refuse to reply any query without any outcomes of any type.

I take into account that I can withdraw permission to use records from my interview within two weeks after the interview, in which case the material may be deleted.

I have had the cause and nature of the study defined to me in writing and I have had the opportunity to invite questions about the study.

Participants Signature: _________________________________

Date of Participants Signature: ___________________________

Researcher’s Signature: __________________________________

Date of Researcher’s Signature: ____________________________

12.4.      APPENDIX D (Interviews)

12.5.      APPENDIX E (Survey)

 

 

 

Cite This Work

To export a reference to this article please select a referencing stye below:

Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.

Related Services

View all

DMCA / Removal Request

If you are the original writer of this essay and no longer wish to have your work published on UKEssays.com then please: