#carlifornia | Explore Tumblr Posts and Blogs

l-in-c-future · 2 months

Text

How AI big data policing can reinforce and amplies bias and prejudice?

In the book Partial Truth: How fractions distort our thinking, Zimring explained how the rise of big data policing could amplify and reinforce bias and prejudices, even in a democractic country. Zimring's arguments were referenced from Andrew Gutherie Ferguson's The Rise of Big Data Policing: Surveillance, Race and the Future of Law Enforcement (NY University Press, 2017); Carlifornia State Auditor's report on the GalGang Criminal Intelligence System Report in August 2016, Chabria, et al. "California Bar Police from using LAPD Records in Gang Databasem Critics Want It Axed, Los Angeles Times, 14 July 2020 and the Editorial Board, "Who will kill or be killed in violence-plagued Chicago? The Alogrithm knows," Chicago Tribune, May 10, 2016.

What are the biases and prejudices AI big data can amplify that endanger enforcement discrimination that can endanger democracy? 1. Big data is NOT as those who advocate it as reducing human bias and prejudice because the data population where the data was compiled could be contaminated at multiple sources. Sources contamination and distortion would lead to bias alogrithimic results.

2. The ways how alogrithims are applied could be be biased. In another words, alogrithms methodologies could sustain bias and prejudices and even amplify them in subtle ways that the users of big data systems are not aware of the hidden biases.

3. It is a worry trend that innocent unrelated people can be falsely read flagged by the rise of big data big AI policing and surveillance. Galgang had wrongly flagged 42 babies as gang members because they happened to be RANDOMLY and UNCORRELATEDLY met two red flags criterion. Worse, people could be wrongly red flagged and wrongly ASSUMED to be related to gang members simply because they lived in areas where gang members HAPPENED to be living in the same area.

4. Big data policing can distort police's exercise of probable cause and reasonable suspicions in searching people's properties and making arrests because of the addresses they live.

5. The ways how the criminals/crimes 'fever billbord' generated by AI's alogrithm produced big data can influence numerous biases such as: distortion of the numerators and denominators of the data in the fractions in percentages; frequency; and proportional references to the data; attributable substitution; availability heuristic; confirmation biase and cherry-picking selective bias to reinforce existing assumptions. e.g. even two young men committed the same crime driven by different by different motives can receive different treatments by the police and the judicial system and whether the two people will leave a criminal record in the 'fever billboard' to become an automatic easily flagged person in the persons of interests billboard. SIMPLY because a person's address locates in a poor area WILL make him more likely to be stopped search by the police because that area is ALREADY red flagged by the big data as crime prone hotspots. (I experience similar traffic fare checks when I move suburbs. The frequency of the buses being stopped to check by a Traffic Officer is up to few times a week compared to the suburb I lived in the past and ALL other affluent suburbs I been to where I RARELY met ONCE even anybody check the fares every 10 yrs!) On the other hand, the same age young person lives at an affluent area has much lower chance being stopped search. So if the probabilities that BOTH of them actually commit crimes, in Ferguson's case cited as selling drugs to earn extra money driven by different motivations, the person living at poorer districts have MUCH higher chance to be stopped searched because the big data flagged his area as high risk where there will be more police presence. Under such situation, the person from poorer suburb will have higher chance to be caught for the same crime being committed as the other person who lives at an affluent area. Furtherdown even BOTH of them are caught, what lay before them can be different. The rich young person may be more accessible to better legal supports to defend against the crime, more ampathy from schools and the community. As a result, there is a higher chance that there will no criminal record left in the system. But the poor person won't have the same resources to defend himself. He only has one outcome whether he admit guilty or not-being convicted. The person will lost the job or evicted from school. ONCE he is convicted, he WILL leave behind a criminal record in the big data system. Even he is on bail or he has ANY on bail history, the alogrithm of choosing what are the 'high crime risks' criterion WILL put him back to the 'fever billboard'. And because he is reinforceably being red flagged by various flagging criterion, this person will have more chance to be stopped search. HOW ABOUT the reality is that a poor person who struggles to even earn enough to keep a living and therefore more prone to engage in criminal activities compare to affluent people who engage in criminal activities BECAUSE THEY WANT MORE MONEY to the ALREADY more money? How about the reality is that BOTH young people or adults from an affluent backgrond and a poor background have similar probabilities or chances to commit crimes? And even the crimes committed by people from affluent background are more severe than the crimes committed by those who come from poor background? The ways how the alogrithms or the designers of the big data's alogrithm's calculations can be overly generic or with hidden biase that are unbeknown to the users and designers of the big data. Worse, it is the designers DELIBERATELY put in bias and prejudice because of systematic injustice in enforcing the laws.

Big data policing can lead to enforcement injustice even in democracy, let alone in authoritarian and autocratic regimes, the big data policing under the state's BIG BROTHER SURVEILLANCE are common suppressive tools used to arrest and prosecute whoever the authorities want to silence and persecute.

Basing on the GalGang's 'red flag' criterion, I can elaborate how the Big Data Policing had caused police terror in HK during the 2019-2020 protests and after the PRC's National Security Law and Article 23 were introduced in HK. From what the police there had enforced how the searched, arrested, detained people, it is not difficult to deduce what are the big data policing flagging criterion used against people who the authority THINKS AND ASSUMES they are 'pro-democratic' "dissidents.

Those who admit to be members or leaders of any pro-democratic political parties or civil groups.

Those who are found or ASSUMED to be related or connected to pro-democratic public figures.

Those who are known to be members or leaders of any pro-democratic political parties or civil groups.

Those who are ACCUSED by China, the HK authority as 'dissidents';

Those who are SEEN 'frequently' appearring in protests risk places at critical political dates such as 8964 or 1st July namely Causeway Bay and Victoria Park of Causeway Bay. Police deployed heavy presence ANNUALLY. They surveilled NORMAL ORDINARY people who turn up there and are ready to arrest.

Being SEEN as wearing clothes or holding items that labelling themselves as pro-democratic or 'dissidents' who meet criteria above 5. i.e. Whever the police SEES somebody wearing a black shirt, a yellow shirt holding a candle, a bunch of flowers, ANYTHING showing 8964 at Causeway Bay WILL BE SURELY arrested.

Detention, categorization and interrogations according to whatever 'national security' crimes imposed by the laws imposed by China to HK.

Being SEEN as showing symbols, signals or hand gestures that let the authority THINKS that the individual is associated with advocacy of democracy or 8964 or ANYTHING China doesn't like to see or hear. This is one of the most common reason why an individual is arrested.

According to reliable sources of reporting or sources of reporting against a person being accused as 'pro-democracy' and THEREFORE AUTOMATICALLY treated as 'dissident' by the authority.

From UNVERIFIED and UNSUBSTANTIATED sources of allegations that a person or entity is 'pro-democratic', 'seditious', 'opposing' the authority or 'criticising' the authority. This is the MOST WIDELY AND VAGUELY source how a person is flagged in the big data BIG BROTHER's policing system through which how any person and entity is persecuted, arrested and prosecuted. Underlying the big data BIG BROTHER survelliance AI is the political ideologies and the VAGUE extensions of interpretations of the ideologies DELIBERATELY built in by the police system and the entire political system.

As a result, in a democracy, the prosecutors and jurors can fall into the trap of the prosecutor's fallacy like in the case of People Vs Collins. But in an authoritarian system where political crimes are trialed WITHOUT jurors and judicial independence judges, the trials will be conducted ALL in accordance with both the police and the prosecutors' fallacies. The verdicts had been long nailed before even the trials were conducted because the 'crimes' were already decided from the MOMENT a person is flagged in the big data policing system SOON AS they are politically labelled by China as 'problematic'.

Zimrang pointed out one IMPORTANT thing: big data is NOT the antidote of human's own prejudices and biases. The ways how people design and use big data alogrithms can worsen and amplify PRE-EXISTING prejudices and biases. Such biases don't even require human beings to cautiously and cognitively aware of, it only requires specific ways to code and programming the formulae in applications. And as long as whatever coded formulae are built into the alogrithms of the AIs, any silicon based computer processors will be equipped with such biases. Unfortunately, this is EXACTLY the current situations. So what is the implication to avoid big data to worsen and amplify prejudices and biases? I think the following are minimum essentials:

ALL sources information and data from which the machine learning alogrithms used as bases for deepn learnings MUST be made open to be counter examined and scruntinised by both the public.

2. Big data generated by AI, especially those produced by generative AIs, has to be flagged as AI generated contents.

3. The alogrithms and how they are built including the formulae how the alogrithms work have to be transparent and independently auditable, questionable and challengeable, especially in areas of applications that have significant public interests attached, such as enforcement, judicial uses, public health, public transports, critical infrastructures, etc....

4. A healthy check and balance to AI big data is not just having abundant of people who know how to made AIs or designs AI system, more fundamentally, it is imperative for the public to have the abilities to question, scruntinise, challenge, query the big data and how they are compiled. A lot of these skills are back to ABC: do we know and how much do we understand human prejudices and biases embedded in our thinking? How much can we detect sources of data contamination in big data machine learning? The public must be given a choice of opt out without subjected to political, economic, and social discriminations as a result of the choices to opt out big data and AI. The last part is most fundamental principle of a democracy. Choices must be given to the affected people without them being subjected to the disadvantages of imbalance of powers between individuals and the public versus the enormous supreme powers of those VERY FEW big technologies companies that develop and design big data.

#Big data #big data policing #AI #state surveillance #prejudice #bias

0 notes

robertnelson2-blog · 3 months

Text

Board Certified Radiologist Germany

Dr. Samuel Kobba embarked on his illustrious medical journey at the University of Bonn Medical School in 2001, where he laid the foundation for a career that would span various aspects of the medical field, especially in research and radiology. His academic pursuit in medicine was rigorous and comprehensive, culminating in 2006 when he was on the cusp of completing his medical degree, pending his final exams. However, his passion for research led him to the University of California, Davis, where he undertook a doctorate thesis at the Genome and Biomedical Sciences Facility under the expert guidance of Professor Knowlton. During this period, Dr. Kobba delved into the intricate process of isolating cardiac myocytes—a procedure that demanded nearly six hours of meticulous work, followed by a series of experiments utilizing heat shock factor and TNF. These experiments employed cutting-edge techniques such as Western blot and ELISA (enzyme-linked immunosorbent assay), both of which are paramount in the separation, identification, and detection of proteins and antigens in biological samples, through highly specific antibody-antigen interactions. His groundbreaking work in this field not only showcased his commitment to medical research but also earned him the title of Medical Research Scholar of the Year. Board Certified Radiologist Germany

Dr. Kobba’s doctoral thesis, “Heat Shock Paradox: Role of Heat Shock Factor,” was a testament to his innovative research and was subsequently published in a renowned medical journal, marking a significant milestone in his career. Furthermore, during his time at the University of California Davis, he immersed himself in clinical rotations at the University of California Davis Medical Center in Sacramento, working under the tutelage of Dr. Matthew Bobinski, Dr. Arthur Brooks Dublin, Dr. Rosalie Hagge, and Dr. Adam Greenspan, where he completed his courses with honors.

Upon his return to Germany, Dr. Kobba successfully completed the State exams in 2009, which allowed him to begin to practice. He commenced his professional journey as a resident physician at the University Clinic in Bonn, where he continued to engage in research. Throughout his journey, he was mentored by luminaries such as Professor Baumgarten, Professor Knüffeman, and PD Dr. Se-Chan Kim, who played pivotal roles in shaping his career. He extends his heartfelt gratitude to these mentors, especially the late Professor H. Shimassek, his personal mentor, and Dr. Seitz, who supported him through challenging times. His brother, Dr. Joseph Kobba, his cousin and wife, Mr. and Mrs. Strasser-King, in Sacramento, Carlifornia, also deserve special mention for their unwavering support throughout his academic and professional journey.

Dr. Kobbas career path led him to Berlin in 2004, where he continued his training at institutions, like Vivantes Klinikum and Charite. After completing his residency he worked as a fellow at Vivantes Klinikum until 2008. Then transitioned to a role as an attending physician at Saint Augustinus Krankenhaus in Düren. During this time his expertise in radiology flourished, eventually leading him to a consultancy position in 2019. In collaboration with Professor Brassel and Professor Sommer he contributed to a chapter titled “Vene of Galen Aneurysmal Malformations” in the book “Pineal Region Lesions; Management Strategies and Controversial Issues.”

His career took a turn when he joined Radio-log, a network of practices in Bavaria. However the outbreak of the COVID 19 pandemic prompted him to move to blikk Radiologie (now known as evidia) which has than 75 locations throughout Germany. Additionally he made contributions to Klinikum Aschaffenburg and Klinikum Darmstadt as a consultant. Dr. Kobbas broad experience in radiology spans areas such as neuroradiology, musculoskeletal radiology, cardiovascular radiology( cardiac computed tomograms and cardiac magnetic resonance imaging) , ultrasound examinations, Ultrasound and CT guided procedures and particularly MRI diagnostics. His work in Doppler ultrasound technology and his emphasis on diagnosing and treating strokes demonstrate his commitment, to advancing knowledge and improving patient care. Consultant Radiologist Frankfurt

The Radiology Department, at Klinikum Darmstadt has provided me with an opportunity to enhance my skills in radiology significantly. Although the workload was initially daunting, I saw it as a chance to grow than shy away from it. Stepping out of your comfort zone is essential for development. I had the chance to perform procedures ranging from computed tomography guided draining of abdomenal abscesses to intricate lung, liver and bone biopsies from different parts of the body with a focus on the back (vertebral column), particularly in cases like spondylodiscitis. My proficiency in conducting and interpreting cardiac computed tomograms (cardiac CTs) and cardiac magnetic resonance imaging (cardiac MRIs) has improved tremendously. Being part of Praxis Radiolog, in Bavaria based in Passau and Klinikum Aschaffenburg made me feel like I belonged to an caring community.

0 notes

techcreecamper · 11 months

Text

Bullimania 6 von 10

So ändert sich das im Laufe der Zeit. Früher hat man einfach mehr Wert auf das Äußere gelegt. Mal ehrlich?

Während man wohl früher mehr Wert auf das Äußeree legte, so hat sich das im Laufe der zeit ziemlich gewandelt. Heute sind es eben die inneren Werte worauf es den leuten ankommt. Und in der tat könnte man es so formulieren. Die ersten Bullis waren entsprechend ihrer zeit viel verspielter und detailreicher im Außen-Look. Allerdings war davon dann im Innenraum der Fahrzeuge nicht mehr viel zu merken. Da blieb der Schnick-Schnack eher aus.

Im Laufe der zeit veränderte es sich. Der Bulli passte sich dem Zeitgeschmack an und wurde außen eher schlichter, bekam aber jede menge Komfort und Technik im inneren.Da hatte sich auch die Zielgruppe verändert. Doch es gab den Bulli stets auch in schlichten Versionen bis hin zu sehr luxuriösen und umfangreichen Ausstattungen. Man darf also nicht Gefahr laufen Apfel mit Birnen zu vergleichen.

Mit der Bezeichnung Multivan passte das eigentlich ganz gut, auch wenn VW das ein wenig anders einsortierte. Doch nur wenige Volkswagen Typen waren so multifunktional und damit so flexibel einsetzbar. Aber ein Bulli von 1966 war gegenüber den Bussen von heute schon ein andere Hausnummer. Jedenfalls was das Exterieur, also sein Erscheinungsbild angeht.

Zweifarbenlackierungen gibt es auch heute noch und sie finden heute wieder, wie schon damals, viele Anhänger. Also nicht als Zugfahrzeug, ich meine natürlich Käufer, die sich dafür begeistern. Dabei dürften Kunden von heute weniger den Bezug zu einem Bulli von damals herstellen, sondern vielmehr sich daran erfreuen, dass man einen Bulli eben nicht nur in Uni-Farbe bekommen kann. Grundsätzlich findet man das aber eher bei Sondermodellen, gerade beim VW Carlifornia Ocean kennen wir solche zweifarbigen Lackierungen.

Früher war mehr Lametta, ist also keine Aussage, welche genaugenommen der Wahrheit entspricht. Es zeigt eher wie Menschen dazu neigen reumütig und sehnsuchtsvoll sich an alte Zeiten erinnern zu wollen. Früher war alles besser. Ne, nicht wirklich. Und wir haben das schon immer so gemacht. Glatt gelogen. Also raus die Waschschüsseln mit kaltem Wasser für eure Bäder und um auf's Klo zu gehen, dann über den Hof und sich auf's Blumsklo setzen. An der Stelle kommt dann nicht mehr viel Schwärmerei in Nostagie auf.

Und so ein wenig ist es ja auch beim Bulli. Die alten mögen uns ja gefallen und ins Schwärrmen bringen und wohl dem/der der/die einen besitzen. Aber mal ganz ehrlich, für den t#glichen Einsatz und auch auf langen Strecken? Naja, da bin ich doch froh, dass so ein Bulli von heute mehr Komfort bietet. Bei aller Liebe könnten wir doch sicher auch Freude an einem alten T1/T2 zeigen und dennoch weniger polemisch sein. Oder etwa nicht?

0 notes