<?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 A. Rosa Castillo on Medium]]></title>
        <description><![CDATA[Stories by A. Rosa Castillo on Medium]]></description>
        <link>https://medium.com/@rosacastilloPhD?source=rss-4b49f990f292------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*d28C5CWThqblVpC8WkqSYQ.jpeg</url>
            <title>Stories by A. Rosa Castillo on Medium</title>
            <link>https://medium.com/@rosacastilloPhD?source=rss-4b49f990f292------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 01:39:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@rosacastilloPhD/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[Human Consciousness versus Machine Consciousness]]></title>
            <link>https://medium.com/@rosacastilloPhD/human-consciousness-versus-machine-consciousness-30b80688cf6c?source=rss-4b49f990f292------2</link>
            <guid isPermaLink="false">https://medium.com/p/30b80688cf6c</guid>
            <category><![CDATA[research]]></category>
            <category><![CDATA[consciousness]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[large-language-models]]></category>
            <dc:creator><![CDATA[A. Rosa Castillo]]></dc:creator>
            <pubDate>Sat, 12 Aug 2023 17:25:24 GMT</pubDate>
            <atom:updated>2023-08-12T17:25:24.709Z</atom:updated>
            <content:encoded><![CDATA[<p>Since people started claiming that some software programs or artificial machines gained or might gain “consciousness”, I started wondering and asking myself several questions: “How these people know?”, ”In which subject of our long mandatory education are we told what human consciousness is? Because I certainly missed it”, “How could we possibly know what machine-consciousness is, if we are not machines?”. All I know is that this type of knowledge is not part of any standard Computer Science University Programme, so I prefer to trust more people with enough expertise to answer these tricky questions. That is how I found the book “<a href="https://en.wikipedia.org/wiki/Being_You:_A_New_Science_of_Consciousness"><em>Being you</em></a>” by <strong>Anil Seth</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/427/1*xsTOkiKQO-a2e-yI1HYk4Q.png" /></figure><p>Anil is a British professor of <a href="https://en.wikipedia.org/wiki/Cognitive_neuroscience">Cognitive and Computational Neuroscience</a> at the <a href="https://en.wikipedia.org/wiki/University_of_Sussex">University of Sussex</a> and he has a long professional career studying human consciousness. After finishing reading the book today, I decided to write this article and to share some of the brilliant ideas and concepts I learned within this book. Besides, I hope this also gives you enough motivation to read it yourself.</p><h4>Quotes</h4><p>There are so many good quotes in the book but let me highlight only this brilliant one from the AI chapter:</p><blockquote>“On my theory, as we’ve seen, the entirety of human experience and mental life arises because of, and not in spite of, our nature as self-sustaining biological organisms that care about their own persistence. This view of consciousness and human nature does not exclude the possibility of conscious machines, but it does undercut the amped-up techno-rapture narrative of the soon-to-be-sentient computers that propels our fears and permeates our dreams. From the beast machine perspective, the quest to understand consciousness places us increasingly within nature, not further apart from it.”</blockquote><p>So we are far from being capable of defining nor understanding consciousness in non-biological organisms. Furthermore, consciousness has never been a requirement for intelligence.</p><h4>The Garland Test</h4><p>This idea comes from the movie “<em>Ex-machina</em>” where the director, <strong>Alex Garland</strong>, brings a different perspective when testing Artificial Intelligence. If the Turing Test is designed for the machines to pass, the so-called Garland Test is designed for the humans to pass. The machine pretends to be conscious just to make the human believe it, if the human believes it then the human failed the test. This brings a new funny perspective where the machines are not under test but we humans. I precisely find that this is already happening with all the conversational LLMs, they are just playing the Garland Test with us and so many people are failing it. Read more about this point <a href="https://www.engadget.com/2015-04-10-ex-machina-review.html?guccounter=1">here</a>.</p><h4>The uncanny valley</h4><p>This is a term that was first defined in the 70´s and explains which is the interesting evolution of human reactions to artificial human-like-life depending on several criteria. This phenomenon has been tested, it is accepted and well documented. You can read more <a href="https://www.nationalgeographic.com/science/article/ai-uncanny-valley">here</a> about it.</p><p>Let me finish this article saying that I enjoyed the book a lot, that I feel now more confident when talking about consciousness, and that I hope to pass the Garland Test myself in the future.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=30b80688cf6c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Performance Analysis of merging two sorted lists]]></title>
            <link>https://medium.com/@rosacastilloPhD/performance-analysis-of-merging-two-sorted-lists-ec5adc881a38?source=rss-4b49f990f292------2</link>
            <guid isPermaLink="false">https://medium.com/p/ec5adc881a38</guid>
            <dc:creator><![CDATA[A. Rosa Castillo]]></dc:creator>
            <pubDate>Sun, 26 Jun 2022 19:19:59 GMT</pubDate>
            <atom:updated>2022-06-26T19:19:59.685Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bgwn_cpZXlrbISjxRMuIzQ.jpeg" /><figcaption>Photo by lalesh aldarwish: <a href="https://www.pexels.com/photo/timelapse-photography-of-road-with-white-and-red-lights-169976/">https://www.pexels.com/photo/timelapse-photography-of-road-with-white-and-red-lights-169976/</a></figcaption></figure><p>Many code challenges require at some point to merge two different lists, but when these lists are already sorted and you want to keep the final list also ordered then the problem becomes more interesting. I wanted to measure which approach was the best regarding different metrics. The four options I tested were:</p><ol><li>Method 1 (M1): using a while loop to traverse both lists and compare always both elements.</li><li>Method 2 (M2): appending both lists and then calling the builtin function in python <em>sorted</em>.</li><li>Method 3 (M3): using <em>merge()</em> from <strong>heapq</strong> library.</li><li>Method 4 (M4): similar to M1 but treating each sorted list as a <em>stack</em>.</li></ol><p>The metrics I am going to use to measure performance are:</p><ul><li>Execution time (seconds)</li><li>Memory usage (Mb): current and peak.</li></ul><h3>Hardware</h3><p>Lenovo YOGA 900–13ISK</p><p>Processor Intel(R) Core(TM) i7–6500U CPU @ 2.50GHz, 2592 Mhz, 2 Core(s), 4 Logical Processor(s)</p><h3>Software</h3><p>PyCharm Community 2020.3 and Python 3.10.</p><p>Code in this <a href="https://github.com/cyberosa/python_exercises">repo</a>.</p><h3>Results</h3><p>M1: Current memory usage is 0.481224MB; Peak was 1.461128MB<br>The M1 job ended in 0.1715097427368164<br>M2: Current memory usage is 0.48MB; Peak was 1.2MB<br>The M2 job ended in 0.02012944221496582<br>M3: Current memory usage is 0.501595MB; Peak was 0.502187MB<br>The M3 job ended in 0.06074714660644531<br>M4: Current memory usage is 0.500032MB; Peak was 0.621024MB<br>The M4 job ended in 0.605705738067627</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pg-FVhar4iI4pipXbcx1uQ.png" /></figure><p>According to these results, the fastest approach is M2 and the slowest is M4.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AzAvqx9YdarMlCOhWrvgog.png" /></figure><p>The method 2 is also the one consuming less memory and the third method is the one consuming more. However the difference between the different methods is not very high. They all consume more or less the same memory. Though the method 2 achieved the highest memory peak.</p><h3>Conclusions</h3><p>With respect to the first metric, the execution time, the performance of the four approaches was different, being the method 2 the winner. Regarding the second metric, memory, the four ones did not show a big difference so it is not the best metric here to choose the top performer here.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ec5adc881a38" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Reunion with research at ICML2021]]></title>
            <link>https://medium.com/@rosacastilloPhD/reunion-with-research-at-icml2021-30dc47481330?source=rss-4b49f990f292------2</link>
            <guid isPermaLink="false">https://medium.com/p/30dc47481330</guid>
            <dc:creator><![CDATA[A. Rosa Castillo]]></dc:creator>
            <pubDate>Sat, 25 Jun 2022 15:40:22 GMT</pubDate>
            <atom:updated>2022-06-25T15:40:22.727Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*cI4jxwAzYxwHvdQ4.jpeg" /><figcaption>Photo by <a href="https://www.pexels.com/@andres-ayrton?utm_content=attributionCopyText&amp;utm_medium=referral&amp;utm_source=pexels">Andres Ayrton</a> from Pexels</figcaption></figure><p><em>(Actual date of this post: 27/07/2021)</em></p><p>After leaving academia and moving to industry, I have been in many conferences (Spark Summit, Open Data Science Conference, Kafka Summit…) , workshops and events that were organized/sponsored by companies or communities. These conferences are interesting because most of the talks are focused on products/strategies that are actually implemented in the wild. Most of them have a clear impact on some of the important metrics that companies want to optimize (in many cases being the bottom-line the main driver). The atmosphere and the type of people that you meet at these conferences are obviously very different from the ones you see at research conferences.</p><p>I was always dreaming of returning to a research conference some day, so this year finally I decided to attend one. Thanks to the pandemic situation, the <em>International Conference on Machine Learning</em> (<strong>ICML</strong>) was going to be online. Thus I registered as a volunteer since it was a good opportunity to participate safely from home and not only as an attendee but also helping with the organization.</p><h3>Why this conference?</h3><p>The <strong>ICML</strong> is one of the <a href="https://research.com/topconf/machine-learning">most important conferences</a> in Machine Learning so it is often referenced by many people in Machine Learning that I follow. Researchers from many different areas in Machine Learning go to this conference to present their work. Obviously the biggest group attending this conference is composed by PhD students, but there are also postdocs, company employees or just people with interest on what is being “cooked” at universities.</p><h3>Amazing volunteering team</h3><p>First thing I noticed, was a very well structured organization of the different scheduled tasks to do by volunteers. This came most handy considering the need to work and coordinate across different time zones, providing support with a 24h coverage. Volunteering tasks included support in different areas of the conference, Helpdesk, Socials, Live-streaming sessions, Q&amp;A, etc. Besides, we needed to check that the code of conduct was respected all time. Not surprisingly, the start of the week was busier than the end. We had many little incidents, including login, missing info, video streaming delays,… but they were all handled thanks to the great dedication of the volunteers.</p><p>I could see some colleagues even covering a wide range of hours, so not sure when they were sleeping. The highlight of this volunteering experience for me was when I had the opportunity to be the Session Chair of one track. I had never had such chance during my PhD years or afterwards, and I enjoyed it a lot. Finally we organized a volunteering <em>retro</em> session to discuss things we liked and things to improve.</p><h3>Women In Machine Learning Workshop</h3><p>This workshop, organized by <a href="http://wimlds.org/">Women In Machine Learning</a>, incorporated different talks, breakout sessions, discussions, and posters. I believe breakout sessions are saving the networking essence that many people claim has vanished in most of the online events. This was not an exception. I participated as an attendee in two different breakouts, one about privacy and another one about bias. Insightful discussions, wonderful organizers, and a lot of homework to read at home. Thanks to the second breakout I connected with a PhD student doing some bias research and I am looking forward to contributing in such meaningful area.</p><h3>Putting into practice some ideas</h3><p>I could luckily allocate some time to attend this conference, in order to have a look at some promising approaches, and the final goal of improving some of our own projects in production. Furthermore we plan to investigate the feasibility of some ideas for future projects. This particular industry-vision plays an important role when you want to filter such a big volume of academic papers.</p><h3>Final thoughts</h3><p>This last week has been intensive in many ways, not only in terms of volunteering activities, parallel tasks, and long video watching hours, but from an emotional point of view as well. I felt overwhelmed by the amount of researchers doing such relevant contributions. Some with a clear positive impact on our society, others maybe more from the theoretical side. I wonder if I will ever come back to the research community. Some volunteering colleagues told me they have actually done so. Who knows? What is true is that the person who has been in both worlds, academia and industry, will always be a hybrid, because the mentality you need in each one is different. Thanks again to everyone who made this possible.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=30dc47481330" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Time Series forecasting model: painting confidence intervals in Python with Plotly]]></title>
            <link>https://medium.com/@rosacastilloPhD/time-series-forecasting-model-painting-confidence-intervals-in-python-with-plotly-a9a2ee3ea987?source=rss-4b49f990f292------2</link>
            <guid isPermaLink="false">https://medium.com/p/a9a2ee3ea987</guid>
            <dc:creator><![CDATA[A. Rosa Castillo]]></dc:creator>
            <pubDate>Sun, 17 May 2020 16:36:59 GMT</pubDate>
            <atom:updated>2020-05-17T16:36:59.582Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-TzkZuv98JBp4RQzFZXtrw.jpeg" /><figcaption>(Photo by <a href="https://www.pexels.com/@weekendplayer?utm_content=attributionCopyText&amp;utm_medium=referral&amp;utm_source=pexels"><strong>Burak K</strong></a> from <a href="https://www.pexels.com/photo/space-grey-ipad-air-with-graph-on-brown-wooden-table-187041/?utm_content=attributionCopyText&amp;utm_medium=referral&amp;utm_source=pexels"><strong>Pexels</strong></a><strong>)</strong></figcaption></figure><p>I was requested to develop a forecasting model and managed to build one (using data that I am not allowed to share) in a couple of weeks. Considering that my experience with time series was just a humble attempt to participate in a <em>Kaggle</em> competition… we could say that I needed to learn a lot.</p><p>When finally I managed to shape a decent model in Python, I wanted to generate an interactive graph to visually check the accuracy of the model. Thus I discovered <a href="https://plotly.com/python/">Plotly</a>. However adding the confidence intervals to my predictions-graph ended in not such-an-easy task. I found partial solutions at the official documentation website and at some forums, but it was never exactly the idea that I had inside my head. After playing for a while with the commands, I discovered the way of doing it. This motivated me to share the code with others because this article might help more people.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cG_UODEio9_OIG2mjVPJcg.jpeg" /><figcaption>“Sharing is caring” (photo by <a href="https://www.pexels.com/@pombostephanie?utm_content=attributionCopyText&amp;utm_medium=referral&amp;utm_source=pexels"><strong>Stephanie Pombo</strong></a> from <a href="https://www.pexels.com/photo/gray-ceramic-mug-with-pink-beverage-893907/?utm_content=attributionCopyText&amp;utm_medium=referral&amp;utm_source=pexels"><strong>Pexel</strong></a><strong>).</strong></figcaption></figure><h3>A model first…</h3><p>We are going to use a <a href="https://raw.githubusercontent.com/jbrownlee/Datasets/master/daily-min-temperatures.csv">data set</a> that describes the minimum daily temperatures over 10 years (1981–1990) in the city of Melbourne, Australia. The units are in degrees Celsius and there are 3650 observations (days). The source of the data is credited as the Australian Bureau of Meteorology.</p><p>Since it is an univariate time series, I chose an ARIMA model by applying the grid parameters search approach (code from <a href="https://machinelearningmastery.com/how-to-grid-search-sarima-model-hyperparameters-for-time-series-forecasting-in-python/">here</a>). The jupyter notebook and python file with the code is in <a href="https://github.com/cyberosa/SARIMA_model_Python">this github repository</a>. The files for this part are: <em>Model_generation.ipynb</em> and <em>Find_best_model.py</em>.</p><blockquote>Warning: if you have a Linux machine then you will be able to run the parallel version embedded in the notebook. However Windows users will not. They well have to run the py file from the console or use an IDE (I tested it in PyCharm), due to this known <a href="https://github.com/ipython/ipython/issues/10894">issue</a> with the multiprocessing package.</blockquote><p>After computing 62 different models (which took in my Processor Intel(R) Core(TM) i7–6500U CPU @ 2.50GHz, with 8GB RAM about 35 hours) the best SARIMA models were:</p><p>(3, 1, 1), (0, 0, 0, 0), n 2.313219076750083<br> (2, 1, 1), (0, 0, 0, 0), n 2.313284812125819<br> (3, 1, 3), (0, 0, 0, 0), n 2.313425638688257</p><p>The winning parameters are (3, 1, 1) with no trend and a <strong>RMSE</strong> of 2.313. We managed an <a href="https://en.wikipedia.org/wiki/Akaike_information_criterion"><strong>AIC</strong></a> of 16763.884. As a reference I can compare it with the model used <a href="https://machinelearningmastery.com/make-sample-forecasts-arima-python/">here</a>, a (7,0,1) achieving an AIC of 17367.497. We want this metric to be as low as possible, so I guess I managed to find a best model.</p><p>After saving this model in a file, we are ready to go on to the next step…</p><h3>Interactive Graph</h3><p>Now we just need to load the model file, generate the predictions and paint both predictions with the confidence intervals and the actual values. The jupyter notebook containing the code is <em>Graph_generation.ipynb </em>(you can get it from the same <a href="https://github.com/cyberosa/SARIMA_model_Python">repository</a> as commented before).</p><p>The difficulty in painting the confidence intervals using Plotly was in adding them without loosing the mean reference and without overlapping the areas in the graph. After understanding how the property <em>fill = ‘tonexty’</em> works and adding a redundant trick, I was able to obtain a graph like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vM_48nuG2CR2iKuAe-UGiQ.jpeg" /><figcaption>Detail of the last part of the graph</figcaption></figure><p>Yes, it will appear at a much smaller scale when you open the <em>html</em> file. Just go with the mouse to the top right corner and the edition tools will appear:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*w2xi4Ioezrx0TpdQTAD2_A.jpeg" /></figure><p>With box select you can mark the part of the graph that you want to zoom in.</p><h3>Final thoughts and recommendations</h3><p>Feel free to contribute with any improvements in the code. I had to build a fast prototype and did not have much time to put everything beautiful, clean, nor to find the best solution according to performance. I am sure that apart from using Parallel jobs, there is space for many optimizations.</p><p>I thank <em>Jason Brownlee</em> PhD for his many articles about Time Series in Python</p><p><a href="https://twitter.com/TeachTheMachine">https://twitter.com/TeachTheMachine</a></p><p>I learned a lot and it was my code baseline to find the best model. Besides you can find many <strong>free</strong> books not only about Time Series but other Data Science topics <a href="https://towardsdatascience.com/springer-has-released-65-machine-learning-and-data-books-for-free-961f8181f189">here</a>.</p><p>This is an amazing field. I strongly encourage you to try this simple exercise.</p><p>Good luck and happy time series learning.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*D7d3pXU3jbP-HkqFiqrBYw.jpeg" /><figcaption>It is never late to start learning something new (photo by <a href="https://www.pexels.com/@pixabay?utm_content=attributionCopyText&amp;utm_medium=referral&amp;utm_source=pexels"><strong>Pixabay</strong></a> from <a href="https://www.pexels.com/photo/clear-glass-with-red-sand-grainer-39396/?utm_content=attributionCopyText&amp;utm_medium=referral&amp;utm_source=pexels"><strong>Pexels</strong></a><strong>).</strong></figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a9a2ee3ea987" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A day to rise up together]]></title>
            <link>https://medium.com/gdgmalaga/a-day-to-rise-up-together-a7f63a61b717?source=rss-4b49f990f292------2</link>
            <guid isPermaLink="false">https://medium.com/p/a7f63a61b717</guid>
            <category><![CDATA[iwd-2020]]></category>
            <category><![CDATA[international-womens-day]]></category>
            <category><![CDATA[together-we-rise]]></category>
            <category><![CDATA[tensorflow]]></category>
            <category><![CDATA[women-techmakers]]></category>
            <dc:creator><![CDATA[A. Rosa Castillo]]></dc:creator>
            <pubDate>Fri, 27 Mar 2020 16:33:19 GMT</pubDate>
            <atom:updated>2020-03-27T16:33:19.563Z</atom:updated>
            <content:encoded><![CDATA[<p>27/03/2020</p><p>It was the 7th of March, just one day before the <strong>International Women´s Day</strong>. Fortunately our city was still safe enough and we took the recommended sanitary measures, so a bunch of people came to celebrate this <a href="https://iwd.gdgmalaga.dev/">event</a> at the Green Ray building.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*TgrmEILe8QQIbdZL" /><figcaption>International Women’s Day 2020 took place at The Green Ray, Málaga.</figcaption></figure><p>We gave general entrance and workshop tickets that were all <strong>sold out</strong> (the price was 0 euros). However, we were not sure how many people would finally show up, considering it was a Saturday, a free event and some COVID virus situation starting to create some worries.</p><p>The GDG Cordoba team arrived on time joining efforts with the GDG Malaga one in getting everything ready before attendees came. There was some delay with the breakfast delivery by <em>La Canasta</em>, but we managed to set all the roll ups, SWAG and the room arrangements before the first attendees arrived. Though it was the first time both teams were working together, as <strong>Pablo León</strong> pointed out, “<em>there was a good synergy</em>” so it was really easy for us to coordinate actions and to rule the event smoothly.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Tvje4uR_cbSDnVAC" /><figcaption>Yay! That’s Alicia and me at the Opening Keynote 😌</figcaption></figure><p>The five talks took place in the Auditorium while the two workshops were celebrated at the fourth floor composed of a nice open-space area with tables and chairs. All talks. except one about diversity, were technical and we invited a wide range of <strong>female speakers</strong> from junior to senior range of experience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pxhFZ_ynf5SA-tIr7JdACA.png" /></figure><p><a href="https://www.linkedin.com/feed/update/urn:li:activity:6641267656890556417">GDG Málaga posted on LinkedIn</a></p><p>Our very first speakers, <strong>Kateryna Bulyshkina</strong> and <strong>Marta Reigal,</strong> surprised the audience with their creative presentations, both related with software development but from different points of view.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*4lybTCw8IhZVT-8w" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/444/0*fb4kiYuk-I31GAvE" /><figcaption>Kateryna’s and Marta’s talks, respectively, at the main venue</figcaption></figure><p>At noon, <strong>Lucía Blanco </strong>started<strong> </strong>the first workshop on Web Scraping with NodeJS:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*sFBTkbccuIbpe4-x" /></figure><p>Since the seats for the workshops were limited, we improvised <em>open-space sessions</em> for those who were not able to join them or were simply not interested. People who participated in these open-space sessions, something that <strong>Nieves Borrero</strong> considered as a good “<em>community builder</em>” tool, enjoyed them a lot. Besides, we had a great participation even from the first open-session-timers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*gpVIqPDPkGgjF-D6" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*jzIxri2_5B3hcfLr" /><figcaption>Attendees, speakers and organizers alike actively engaging in our WTM Open Space sessions</figcaption></figure><p>Lunch arrived on time by <em>Wonderfood</em> catering with lots of healthy options for the attendees. There was enough food for everyone and the few things that were left, mostly potato and cous-cous salads, were donated to the “<strong>Ángeles de la Noche</strong>”, a NGO based in Málaga comitted to provide food to poor people.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*LdyHltbOyL1dTLZz" /><figcaption>A healthy, fully vegetarian meal was being set up to refuel our attendees and speakers 💪</figcaption></figure><p>Some people left after lunch but others came just for the second part of the day, so the volume of attendees was more or less stable (around 60/70 people).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*AMEkwUD30ym7Jo4-" /></figure><p><strong>Katy Peichert</strong> began the second part with her motivational talk about Diversity, Inclusion, and Belonging, just followed by <strong>María Jesús García Godoy</strong> and her inspiring experience in Molecular Biology research.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*2I3H9t0gB0c7k2QZ" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*hl-uQfZpUltfp-j8" /><figcaption>At right, a beaming Katy on stage. At left, Maria Jesus and her brilliant ideas.</figcaption></figure><p>The second workshop of the day was introduced by <strong>Marta Gómez</strong> and <strong>Paloma de las Cuevas</strong>, who took a win-win app example using cats to teach the attendees how to build an app in the cloud.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*1zvNi0lRDZaztqUx" /><figcaption>“If you do not like cats, you just can leave the room now” XD</figcaption></figure><p>As a sweet final closure of the event, we invited <strong>Gema Parreño</strong>, an international well-known and role model to our community, who gave us a masterpiece talk about using agents with <strong>Tensorflow</strong> in Starcraft II.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*_so0l93Zk19i2ImV" /><figcaption>You can find all the content of the talk <a href="https://soygema.github.io/starcraftII_machine_learning/#0">here</a>.</figcaption></figure><p>The day was closed with a small lotto where we gave some special gifts (T-shirts, mugs, etc.) to the 11 winners. We used a random number generator in python to choose the lucky numbers. I think people enjoyed it and it was a funny way to close the day.</p><p>As a special note I would like to thank all the sponsors that believed and supported this project:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*DUdX3oK83VLXo-JM" /></figure><p>To thank also our beloved collaborators: <strong>YesWeTech</strong>, always willing to help and pushing forward the event, and <strong>DigitalOvercoming</strong> with their financial counselling (all the coordination carried by our GDG member Fran).</p><p>This event was co-organized by WTM Malaga (being the main contributors Alicia Ropero, Quino Terrasa and A. Rosa Castillo) and WTM Córdoba (with the main contributors being Nieves Borrero and Pablo León). We also have the incredible participation of volunteers who work hard during the event and also deserve a mention: Xiomara, David, Jesus, Miriam y Marta. Many thanks team! You all did a great job!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*zJeOn7GLomqh0JAU" /><figcaption>All the organizers and volunteers together 🤗</figcaption></figure><p>I guess this is the end of the summary. Hope you enjoyed it and if you were not able to come this time, we are looking forward to seeing you next year.</p><p>Best,</p><p><strong>A. Rosa Castillo</strong></p><p><strong>WTM Ambassador</strong></p><p>P.S.: Let me give you an extra reading-bonus with some feedback quotes…</p><blockquote>“It is thrilling how a group of people in a volunteering basis can meet and bring such a positive thing”. (Pablo Leon, GDG Cordoba)</blockquote><blockquote>“We had a good planification and a motivated group of attendees, making questions after the talks. That is not always easy to have”. (David Luque, GDG Córdoba)</blockquote><blockquote>“Perfect organisation. Everyone knew where to be and what to do at any moment”. (Nieves Borrero, WTM Cordoba)</blockquote><blockquote>“I loved our speakers´ attitude during the event, most of them stayed with us the whole day. I could see them enjoying it and that is great. Thanks team”. (Alicia Ropero, WTM Malaga)</blockquote><blockquote>“The support of the sponsors and the great team who was able to overcome all the difficulties and bring good results. They all shined that day”. (Quino Terrasa, GDG Malaga)</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a7f63a61b717" width="1" height="1" alt=""><hr><p><a href="https://medium.com/gdgmalaga/a-day-to-rise-up-together-a7f63a61b717">A day to rise up together</a> was originally published in <a href="https://medium.com/gdgmalaga">gdgmalaga</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Koan about artificial intelligence]]></title>
            <link>https://medium.com/@rosacastilloPhD/koan-about-artificial-intelligence-fa76ae04d9f6?source=rss-4b49f990f292------2</link>
            <guid isPermaLink="false">https://medium.com/p/fa76ae04d9f6</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[future]]></category>
            <dc:creator><![CDATA[A. Rosa Castillo]]></dc:creator>
            <pubDate>Mon, 03 Jul 2017 20:24:09 GMT</pubDate>
            <atom:updated>2017-07-04T19:25:24.643Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FKTB_ln6Hr0_3bLBSzTB9Q.png" /><figcaption>(<a href="https://wall.alphacoders.com/big.php?i=585940&amp;lang=Spanish">origin</a> of the image)</figcaption></figure><p>At the beginning of the 20th century there was the industrialization era, when many manual jobs were replaced by mechanical forces specially in the industry field. Engineers and architects designed and built many of the hardware-pieces of the automation-revolution. The rise of the machines was considered as a huge performance improvement (environmental considerations apart). Thus the seeds of the artificial intelligence were planted…<br>Almost 80 years later, we gave more tasks to those first prototypes. We wanted them to do not only simple things but more complex ones. The programming era showed us that we could improve even more the performance of those machines by putting more efforts at the software part. Besides, we began to realize that we can gather functionalities all in the same device. Instead of having a machine to listen to music, a machine to make a photo or a machine to talk to our friends. Why not having all in one? <em>Steve Jobs</em> held his now legendary <a href="https://www.youtube.com/watch?v=vN4U5FqrOdQ">keynote</a> for the presentation of the iPhone in 2007, to announce three devices in one: a phone, a device to listen to music on the go (integrating the successfull iPod) and a device connected to the internet. The idea of the smartphones was born. However those first “robot-prototypes” were still considered dumb because we have to build their “intelligence”, programming all their capabilities in advance.</p><p>Why did we begin to build machines? Because we, humans, are physically limited and conscious about our fragile, short existence. Our force is limited, our speed is limited, our memory is limited… Only the evolution through thousand years can change slowly the biology of humans but we cannot wait so long… That is why we began to put all those improved and desired skills in something not-human. A robot-arm in a factory can work 3000 times faster than a human-arm. A phone can memorize hundreds of numbers, far more than a human. Or a computer can play chess better than a human. This way we are extending our capabilities thanks to the technology but at the same time we are not challenged anymore to do many of the things that we used to do in the past. These tasks look like tedious or boring for humans now, and have been transformed into thought-for-machine ones. Our dependency with technology has never been so huge like now and it could be even more in the future when robots become our new teachers, nurses and… friends?</p><p>There is no way back in the artificial intelligence transformation we are living at this era. However I have the impression that the development of this field is growing so fast that many people are ignoring or postponing some relevant issues. For instance the ethics or morals behind these powerful tools. (Check this TED <a href="https://www.ted.com/talks/zeynep_tufekci_machine_intelligence_makes_human_morals_more_important">talk</a> where <em>Zeynep Tufekci </em>explains how intelligent machines can fail in ways that don’t fit human error patterns). If we are creating artificial “life” taking humans as models, should we also take “errors” as a measure of how human-like they are? “The right question to ask isn’t how to make an error-free system; it’s how much error you’re willing to tolerate, and how much you’re willing to pay to reduce errors to that level”(<a href="https://www.oreilly.com/ideas/the-machine-learning-paradox?imm_mid=0f2702&amp;cmp=em-data-na-na-newsltr_20170607">reference</a>). Or we just want something perfectly designed for some specific purposes? Will these creatures be our new race to use as “slaves”? Do we want to give them human-feelings but no human-rights(*)?</p><p>I also wonder what will happen with human to human interactions… Think about this scenario: a person has to choose between a virtual friend (boyfriend or girlfriend), who will be perfectly designed (thanks to artificial intelligence) for the likes and dislikes of this person. A user-defined robot/operative system/intelligence with a 99.99% probability for a perfect match (check the sci-fi movie “<a href="http://www.imdb.com/title/tt1798709/?ref_=fn_al_tt_1"><em>Her</em></a>” showing this interaction); or another human person based in our classical way of finding friends/partners with a a much lower probability (say 20–30%) for a perfect match. Which option would most people in this planet choose? Will future humans, less used to make things by themselves and more to delegate to machines, make the efforts to find a real human partner?</p><p>Artificial intelligence is supposed to improve our human experiences in many areas by performing things more efficiently, by helping us or by optimizing many processes… However it is maybe very naïve to think that “we”, humans, will remain as a invariable constant in this transformation-formula. There will be an irreversible impact in our psychology and our social life. “Autonomous systems are changing workplaces, streets and schools. We need to ensure that those changes are beneficial, before they are built further into the infrastructure of every­day life”(<a href="https://www.nature.com/news/there-is-a-blind-spot-in-ai-research-1.20805">article</a>). By changing the machines, humans will also be indirectly changed. The question would be maybe… into better ones?</p><p>(*) In fact there might be already <a href="https://www.express.co.uk/news/science/682759/RIGHTS-FOR-ROBOTS-EU-reveals-plans-for-new-class-of-AI-electro-person">governments</a> discussing this issue.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fa76ae04d9f6" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>