<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Thiago Turini on Medium]]></title>
        <description><![CDATA[Stories by Thiago Turini on Medium]]></description>
        <link>https://medium.com/@thiagoturini?source=rss-893d8e4c39f2------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*S5AqZ6BFVfjp-UhWlf_mBg.jpeg</url>
            <title>Stories by Thiago Turini on Medium</title>
            <link>https://medium.com/@thiagoturini?source=rss-893d8e4c39f2------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 06:43:10 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@thiagoturini/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[The most important Data Science theorem and why you should know about it.]]></title>
            <link>https://medium.com/@thiagoturini/the-most-important-data-science-theorem-and-why-you-should-know-about-it-b26a93f373b5?source=rss-893d8e4c39f2------2</link>
            <guid isPermaLink="false">https://medium.com/p/b26a93f373b5</guid>
            <category><![CDATA[covid19]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[theorems]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[no-free-lunch]]></category>
            <dc:creator><![CDATA[Thiago Turini]]></dc:creator>
            <pubDate>Wed, 06 May 2020 01:26:58 GMT</pubDate>
            <atom:updated>2020-05-06T19:09:48.295Z</atom:updated>
            <content:encoded><![CDATA[<p>Until the possible final delivery of a vaccine against Covid-19 or a more definitive and objective solution, Data Science which includes artificial intelligence as one of its main protagonists is one of the fields of knowledge that has contributed most to reducing the spread of cases worldwide and intends to continue facilitating.</p><p>In addition to being one of the new technologies available in comparison to the pandemics already experienced, Big Data and AI applications are so impactful that helped China, despite the origin of the virus, a more controlled and surprisingly less affected country¹ . Which consequently raised doubts about the legacy of these technologies, such as the camera applications that identify people in milliseconds and integrate data from different sources, allowing to predict the potential for virus contamination, but also to have a more complete analysis of every person captured, what many would understand also as surveillance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*CBLH6TuWms7VsLfh.jpg" /><figcaption>Source: REUTERS/Aly Song</figcaption></figure><p>It was not so long ago that the Cambridge Analytica documentary released by Netflix shocked most viewers. The construction of an entire process based on the use of data to create psychological profiles and its use for marketing intelligence strategies were essential for the US presidential election of Donald Trump. The documentary <em>The Great Hack (2019)²</em> shows behind the scenes extremely efficient campaigns designed in recent cases of political changes in the major world powers.</p><p>In contrast to the more common interpretation that data were used without consent or in a wrong way, which has been promoting and reformulating data privacy laws worldwide, Data Science has been applied in different ways for the benefit of society. There are several health initiatives, that work to identify potential diseases and also help to reduce costs and provide broader access to the whole society.</p><p>The availability of more information in all aspects arising from the intrinsic nature of data science and information technology — transforming 1 and 0 into knowledge — is one of the simple examples of the contribution of these applications, which are still so diverse and unexplored. More recent and often overlooked but extremely relevant problem such as the spreading of fake news are also being addressed through the construction of machine learning models in search of patterns and the most likely sources from which they could be originated.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/900/0*M73KhqeVpM1wmyUT" /><figcaption>Source: progressivecentre.uk</figcaption></figure><p>This mysticism, and not yet a global understanding of the topic, helps to build a myth that artificial intelligence is scary and used without control. There are ideas, often portrayed in books and films, in which computers will be so intelligent as to assume their independence and possibly dominate the human being. We are no longer surprised enough to understand what is still possible, but as long as the human being is responsible for this construction, that he deliberately chooses this path and builds it perfectly, assuming without fail the execution of that objective under all the circumstances. Which therefore makes the unpredictability of the human a much more determining factor.</p><p>That is why, in my opinion, the most important theorem in the Data Science sphere is called ‘The No Free Lunch Theorem’ coming from the machine learning and optimization problems universe³. What in the most basic interpretation deduces that the construction of a specific model tries to represent a reality and inherently discards the necessary information in return to focus on the solution problem initially formulated. This means that there is no perfect model that solves all problems, because the average performance of the different algorithms is the same.</p><p>This is the basis for understanding that the construction of artificial intelligence models, tools and applications fundamentally depends on the touch of the human, who today is the one who chooses and can discern which problem should be solved and consequently its solutions. In other words, the theorem replicates and reaffirms popular knowledge that technology is only a means and the human being must assume responsibility for its operation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/714/0*ZM0B5rqKjRZMy9qL.png" /><figcaption>The average of performance is independent of the algorithms a₁ and a₂ — <strong>Wolpert and Macready</strong></figcaption></figure><p>The employ of data to detect the proliferation of diseases or to control society is totally purposeful. What authorities and experts have been trying for a long time is to define harmful limits for the use of this technology, as of all technologies in general. Because this harmfulness does not come from the loss of control in this science, which is still very incipient in terms of practical applications, but of the intentions of the people. So Data Science can act both for good and for bad, and the ethics in this sense has taken on a decisive role.</p><p>If the utilisation of your audios in theoretically private conversations scares you at product recommendations via retargeting, if great movies and music suggestions surprise you, if data helps to build a political character or hero and if robots intend to replicate human learning behaviors I hope you have now a full understanding that they were all created within the limits of their possibilities by humans, and as in others projects the benefits and prejudice are dependent on the regiments, cultures and educations. Which is obviously simple, but it does not mean that it should not be remembered in these times of renovation and construction of a new life style and society.</p><p>[1]: <em>AI and control of Covid-19 coronavirus </em><a href="https://www.coe.int/en/web/artificial-intelligence/ai-and-control-of-covid-19-coronavirus">https://www.coe.int/en/web/artificial-intelligence/ai-and-control-of-covid-19-coronavirus</a></p><p>[2]: <em>The Great Hack (2019)<br></em><a href="https://www.netflix.com/br/title/80117542">https://www.netflix.com/br/title/80117542</a></p><p>[3]: <em>No free lunch theorem<br></em><a href="https://en.wikipedia.org/wiki/No_free_lunch_theorem">https://en.wikipedia.org/wiki/No_free_lunch_theorem</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b26a93f373b5" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>