Close
today i learned (til)

Artificial Intelligence

November 17, 2017

 

According to ‘Wikipedia’, Artificial Intelligence (AI) is defined as the study of ‘intelligent agents’. AI can replace the human identification, recognition, analysis, decision-making and other functions. When you say a word, the machine can recognize the words, and understand the meaning of your words. AI has a history of 61 years first starting from the Dartmouth meeting (the longest continuous bilateral dialogue between American and Soviet representatives) in 1956. During the historical development of AI, there are obvious ups and downs. At least, we have been through two ‘winters’ of AI. Allen Newell and his students win the game checkers, solving word problems in algebra and prove logical theorems. However, later they found that they fail to recognize the difficulty of some remaining tasks. Luckily, according to researchers, we have reached the ‘gold age of Artificial Intelligence nowadays.

FullSizeRender 3

 

Artificial Intelligence has broken the records in multiple fields. In 2016, the latest version of voice-identifier developed by Microsoft lab has reached an error rate of 5.9%. This is the first time when artificial intelligence has the same voice identifying error rate as that of a human being. In the image identifier competition-ImageNet, which uses computers to identify the 1000 kinds of images, ResNet from Microsoft won the championship with an error rate of 3.5% (compare to 5.1% – the error rate of image identifier of a human being).( https://en.wikipedia.org/wiki/Artificial_intelligence#History)

屏幕快照 2017-11-17 下午1.04.12

Besides voice and image, artificial intelligence has a huge improvement in natural language. The new function of Microsoft translator published in December 2016 supports over 50 languages and can achieve simultaneously translates multiple languages spoken or written by different people. In the future, people may encounter others from different countries, but as soon as we use this app, we could hear our mother language.

FullSizeRender 4

 

Voice, image, and language applications are practical. Recently, artificial intelligence has some huge breakthroughs in fields that are complex. We can even use artificial intelligence to play video games. Artificial intelligence has found the quickest ways to get scores in 120 minutes. If we keep training this robot, it will reach the master level of the game in 240 minutes and successfully finds the tricks we cannot find ourselves while playing the game.

(https://www.theverge.com/2015/2/25/8108399/google-ai-deepmind-video-games)

AlphaGo uses a huge amount of data to the self-play process. After a long time of training, it beats the world champion in chess Shishi Li with a score of 4:1. Since it uses self-play process, as long as given enough time and examples, it never stops training itself.( https://qz.com/502325/an-ai-computer-learned-how-to-beat-almost-anyone-at-chess-in-72-hours/) That is why it won masters in chess 60 rounds continuously.

All those new developments in artificial intelligence field make it popular among people from all over the world. However, if we abandon the superficial thoughts we have for artificial intelligence and think deeply of its structure, it is not hard to find the basic problems that are unsolved hidden in artificial intelligence field. This requires all of us who are deeply interested in artificial intelligence to be calm and think about the future development of artificial intelligence.

Nowadays, artificial intelligence can roughly be separated into two parts: deep learning and reinforcement learning. Deep learning uses a very complicated model to imitate the imports and outputs. Reinforcement learning is deciding the best solution through the interaction between learning machine and environment. The reason behind the huge success of deep learning as well as reinforcement learning is that all the models are trained by large databases. However, what if the areas do not have a lot of databases for AI machines to test themselves?

The fact that artificial intelligence needs a lot of databases to train an effective model instead of learning from scratch is a hidden danger in artificial intelligence. Now we are living in the big data world. For images, we can train the models by thousands of images. For voices, there is a huge number of voice data that we can use for training AI. For translators, we can use language samples to train the models. These databases make deep learning possible. However, this is certainly not the perfect solution for all fields. In many areas, it is hard or even impossible to get a similar database. For instance, some rare diseases in the medical field have only a few examples throughout the world. Only if we find the way to overcome the need for big data, can we enable the breakthrough in artificial intelligence technology and broaden the area of it.

The second challenge is about the size of the model and the difficulty of training the models. The internet used for training artificial intelligence always has a large coefficient as of billions.

Facing such a network, there are at least two difficulties: one is that we often refer to the gradient abatement and gradient explosion problems.

This phenomenon described in above steps is what we know as “vanishing gradient problem”.Trying the same approach as mentioned above, but when the observations are totally opposite, it tends to be “exploding gradient problem”.( https://medium.com/@sayondutta/intelligent-signals-unstable-deep-learning-why-and-how-to-solve-them-295dc12a7fb0)

When the deep network has many levels, the output layer and the label between the calculated residual or Loss function are difficult to effectively pass to the bottom. So, in the use of this reverse communication training, the underlying network parameters are usually not very easy to be defined as effective training. They usually have poor performances.

The high coefficient means the storage of the model is so large that it exceeds the storage of a GPU card. Therefore, separate computing and parallel models are required. However, it is time-consuming to get the models.

 

As artificial intelligence is getting increasingly successful in many fields, some people may be concerned that in the future AI may take over the jobs of human beings. However, in my consideration, AI and the work of human being are complementary to each other. Agents are fast, rigorous, precise, explicit and objective. However, they lack common sense and the ability to deal with novel situations. On the other hand, common sense is just the strength of us human beings. Therefore, I believe with the help of AI, more challenges with the rapid development of technology could be solved toward the knowledge economies.

 

Reference:

(1)Tecuci, G Artificial Intelligence March 2012, Vol.4(2), pp.168-180

(2)Tecuci, Gheorghe.Building Intelligent Agents: an apprenticeship multistrategy learning theory, methodology, tool and case studies. Academic Press , 1998

 

 

 

 

 

 

 

 

Tags:
0Likes
Wendy Wei
About Author

Wendy is a second-year ISYE student in Georgia Tech. She loves traveling, listening to concerts and hanging out with her friends outside classes.

You May Also Like

today i learned (til)
Tech is in Bloom for Spring!
today i learned (til)
Why do I sneeze when I step out into the sun?
today i learned (til)
Interstellar’s Movie Depictions: Real or Unreal?
today i learned (til)
Filling in the Gaps: Biases and Pareidolia

Your journey starts here! Let’s talk!

ThemeRex © 2026. All rights reserved.
  Timothy Durham was convicted of rape when a test showed his DNA genotype matched the DNA recovered from the crime scene. Eye witnesses testified that Durham was in a different state at the time of the incident, but he wasn’t exonerated from his 3000 year sentence until follow up DNA analysis showed that the preliminary forensic analysis used in the trial was misinterpreted (Thomas et al.). DNA testing serves an invaluable part in our criminal justice system, but it comes with complications and bias that can lead to wrongful convictions and misinformed juries.  

Background:

Forensic genetics aims to identify the origin of a biological sample. It can be traced back to Karl Landsteiner who discovered human ABO blood groups. With blood groups, suspects could be eliminated based on their blood type. For example, if a suspect had type A blood and there was type B blood left at the scene by the perpetrator, the investigators could rule out the suspect. Unfortunately, the reverse didn’t work. If the suspect and perpetrator both had the same blood type, it can’t be concluded that the suspect was guilty. Blood identifiers evolved into 8 distinct categories which improved the ability to eliminate suspects, but could only conclude that (if the system matched) it might be the perpetrator or this person might be related to the perpetrator (Jobling and Gill). The field evolved in the 1980s with the discovery of Minisatellites, lines of repetitive DNA that don’t encode proteins. Minisatellite patterns in DNA create what we know today as the DNA fingerprint, and it has a much higher accuracy rate than previous methods (Augustyn et al.). Around the same time, a breakthrough for victims of sexual assault was discovered. In the case of assault, bodily fluids of both the victim and the assaulter can get mixed. A process called differential lysis was created which could selectively increase the concentration of sperm in a fluid sample thereby allowing for accurate DNA testing (Jobling and Gill). Forensics evolved to use single-locus probe testing after several years because of its ease of interpretation. It looked at specific minisatellites that had high variability between subjects. This was the first DNA-based method to be used in a criminal investigation. Two murder homicides from several years apart were connected using DNA evidence. The evidence was then used to exonerate an innocent man, and the guilty party was found when he tried to evade mandatory DNA testing for all men in the small town. His guilt was then reinforced when the DNA profiles matched, and he gave a confession (Jobling and Gill). Finding DNA at a crime scene doesn’t guarantee that you’ll find a murderer. There are issues with non-pure samples and limited samples that make DNA work difficult, and accreditation of facilities and new techniques and take time (Jobling and Gill).  

Subjectivity of DNA Interpretation:

In a 2009 publication, the National Academy of Science discusses the possibility for subjective influence in lab based forensics. The interpretation bias happens in complex cases where, for example, there could be a mixture of DNA. A study published in 2011 looked specifically at this issue. A gang rape happened in Georgia, and the mixture in DNA from the case was used to prosecute one suspect after another suspect took a plea bargain to testify against him. With DNA mixtures, there are three possible conclusions: suspect excluded, suspect cannot be excluded, and inconclusive. If the suspect is excluded, it means that there’s little to no possibility that his DNA is part of the sample. If the suspect cannot be excluded, it means that there’s a possibility that his DNA is part of the sample. In the trial, the DNA examiners who knew the context of the case concluded that the suspect being prosecuted could not be excluded. The 2011 study gave the DNA evidence from this case and the DNA profiles of the suspects to 17 forensic examiners. These examiners had no context of the case, and they all worked independently. Of the 17, only 1 came to the same conclusion (suspect cannot be excluded) as the examiners who knew the context of the case. Twelve of the 17 concluded that the suspect who was prosecuted and convicted had no relation to the DNA found in the sample. Not only does this show that subjectivity occurs due to the context of the case, but the 17 examiners came up with different findings based on their past experiences meaning that not all DNA analysis is objective. Some organizations like the Scientific Working Group on DNA Analysis Methods say that quantitative statistical evidence should be presented with every categorical conclusion, but the US doesn’t require this (Dror and Hampikian).

False Positives:

There are two categories that cause wrongful identification in a DNA report: coincidental matches and false positives. Coincidental matches are DNA matches between two unrelated people. The probability of these matches is about 1 to 500,000 for white individuals and about 1 to 350,000 for black individuals. This small probability of coincidental matches makes people feel confident in DNA evidence. False positives, on the other hand, are matches between DNA and test DNA that result from issues in the forensics lab. There have been multiple cases of forensics labs accidentally switching victim and suspected rapist DNA which lead to convictions that were later overturned when the lab realized the error. There are also cases of misreading or misunderstanding a sample. In a case from 1993, a man was convicted of rape because someone interpreted a mixed sample of victim and perpetrator DNA as an individual sample. The conviction was again overturned. To prevent false convictions, retesting DNA is allowed, but it’s not an efficient solution. Many people are pushing for statistical information on the probability of false positives, so that the jurors can make accurate judgements based on DNA evidence (Thomas et al.).  

Conclusion:

DNA analysis is a useful tool for the criminal justice system, but subjectivity can be an issue in forensic testing. To combat wrongful convictions, there are organizations like The Innocence Project who use DNA testing to free the innocent, but reformed forensic testing standards should be put in place to prevent subjectivity and to ensure juries are correctly informed during trials. https://www.innocenceproject.org/    

Citation:

Augustyn , Adam, et al. “DNA Finerprinting.” Encyclopedia Britannica , 26 Apr. 2018, www.britannica.com/science/DNA-fingerprinting#ref153186. Dror, Itiel E., and Greg Hampikian. “Subjectivity and Bias in Forensic DNA Mixture Interpretation.” Science & Justice, vol. 51, no. 4, 2011, pp. 204–208., doi:10.1016/j.scijus.2011.08.004. Jobling, Mark A., and Peter Gill. “Encoded Evidence: DNA in Forensic Analysis.” Nature Reviews: Genetics, vol. 5, no. 10, Oct. 2004, pp. 739–751., doi:10.1038/nrg1455. Thompson, William C, et al. “How the Probability of a False Positive Affects the Value of DNA Evidence.” Journal of Forensic Sciences, vol. 48, no. 1, Jan. 2003, pp. 1–8., doi:10.1520/jfs2001171.