<?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 Chris Dare🔥 on Medium]]></title>
        <description><![CDATA[Stories by Chris Dare🔥 on Medium]]></description>
        <link>https://medium.com/@chrisdare?source=rss-e763f8f5a8f1------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*DL5_5fBu0GjX3oVtb5nHwA.jpeg</url>
            <title>Stories by Chris Dare🔥 on Medium</title>
            <link>https://medium.com/@chrisdare?source=rss-e763f8f5a8f1------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 00:51:25 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@chrisdare/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 Fastest Way to Audit Your Python Dependencies]]></title>
            <link>https://chrisdare.medium.com/the-fastest-way-to-audit-your-python-dependencies-2e1661d91cc3?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/2e1661d91cc3</guid>
            <category><![CDATA[information-security]]></category>
            <category><![CDATA[uv]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[vibe-coding]]></category>
            <category><![CDATA[dependency-management]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Sun, 05 Apr 2026 16:46:03 GMT</pubDate>
            <atom:updated>2026-04-18T07:18:16.168Z</atom:updated>
            <content:encoded><![CDATA[<h3>The Fastest Way to Audit Your Python Dependencies (uv audit)</h3><p>Last week the Astral team released a dependency audit feature in UV. I thought to write about it as my small way of supporting their efforts to improve the pythonic developer experience. Let’s begin.</p><p>If you’re shipping Python code, you’re shipping other people’s code via the 3rd party dependencies you use. That dependency risk has become table-stakes for engineering teams — especially as we adopt AI coding assistants that can spin up new projects and lock in dependencies faster than anyone’s reviewing them.</p><p>Last month alone, two major supply chain attacks proved the point. In March, attackers compromised LiteLLM’s PyPI credentials and published backdoored versions (1.82.7 and 1.82.8) that harvested credentials and SSH keys from 40,000 affected installs. Days later, a North Korean state actor compromised the axios npm account and delivered cross-platform RAT malware to anyone running npm install during a 3-hour window — affecting projects downstream of a library with 100 million weekly downloads. These are things that can be caught with dependency audits in your CI. We have native solutions in rust with cargo audit and npm audit for JS/TS, but it’s been quite the journey for Python.</p><p>Interestingly, most Python developers still skip dependency audits. Sometimes the tooling is either heavy in the case of Snyk, Fortify. Other tools like pip-audit require exporting requirements.txt, and so are disconnected from modern workflows using uv. Some tools, Safety, are pretty good but now paywalled. And some are just bad at doing this for Python — like the OWASP Dependency-Check.</p><h3><strong>Introducing a better alternative</strong></h3><p>Last week, the Astral team shipped uv audit.</p><p><strong>How it works</strong></p><p>For teams already using uv, auditing is now native and free:</p><pre>uv audit --preview</pre><p>Reads uv.lock directly. Exits non-zero on CVEs. Integrates into CI in one line. No tokens, no config, no export step.</p><h4>You can now add dependency audits to CI in two minutes flat</h4><p>After taking it out for a spin — a process that only took 2 minutes to get uv audit integrated into my GHA CI — It immediately flagged some CVEs in my dependencies.</p><pre>Found 2 known vulnerabilities in 88 packages<br><br>anthropic 0.86.0 has 2 known vulnerabilities:<br>- GHSA-q5f5-3gjm-7mfm: Insecure Default File Permissions in Local Filesystem Memory Tool<br>  Fixed in: 0.87.0<br><br>requests 2.32.5 has 1 known vulnerability:<br>- GHSA-gc5v-m9x4-r6x2: Insecure Temp File Reuse<br>  Fixed in: 2.33.0</pre><p>All caught automatically by something that doesn&#39;t require an account, a paid seat/license, or a new tool.</p><p>To get started, just add this to your .github/workflows/ci.yml, open a PR and see the results in action.</p><pre>  dependency-audit:<br>    name: Dependency audit<br>    runs-on: ubuntu-latest<br>    steps:<br>      - uses: actions/checkout@v4<br><br>      - name: Install uv<br>        uses: astral-sh/setup-uv@v5<br>        with:<br>          enable-cache: true<br><br>      - name: Set up Python<br>        run: uv python install 3.11<br><br>      - name: Audit dependencies for vulnerabilities<br>        run: uv audit --preview</pre><p>As more teams adopt Claude, Cursor, or other AI assistants for scaffolding, those tools will lock in dependencies automatically. You need scanning that’s as fast as the codegen. uv audit is, and it understands uv.lock natively — if you’re using uv for deterministic builds, you’re already capturing exact versions.</p><p><strong>Vulnerability databases and roadmap</strong></p><p>uv audit currently queries the <a href="https://osv.dev/">OSV (Open Source Vulnerabilities) database</a>, which aggregates advisories from PyPI/PYSEC, NVD, and other sources. Astral’s roadmap <a href="https://github.com/astral-sh/uv/issues/18506">explicitly plans for multi-database support</a>, with scaffolding already in place for configurable backends. Future versions will likely support PyPI/PYSEC, ecosyste.ms, and other sources directly — but OSV covers the vast majority of Python vulnerabilities today.</p><p>If your team uses uv, you now have no excuse not to scan. If they don&#39;t use uv yet, this is another reason they should.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2e1661d91cc3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI is changing commerce at the decision layer. Here’s what senior leaders need to know]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://chrisdare.medium.com/ai-is-changing-commerce-at-the-decision-layer-heres-what-senior-leaders-need-to-know-9d81ff1bea41?source=rss-e763f8f5a8f1------2"><img src="https://cdn-images-1.medium.com/max/1024/1*kLPiOi6vgt6bQcxpxuKNAg.png" width="1024"></a></p><p class="medium-feed-snippet">AI systems are starting to research, compare, and complete purchases on behalf of customers&#x200A;&#x2014;&#x200A;directly inside search and conversational&#x2026;</p><p class="medium-feed-link"><a href="https://chrisdare.medium.com/ai-is-changing-commerce-at-the-decision-layer-heres-what-senior-leaders-need-to-know-9d81ff1bea41?source=rss-e763f8f5a8f1------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://chrisdare.medium.com/ai-is-changing-commerce-at-the-decision-layer-heres-what-senior-leaders-need-to-know-9d81ff1bea41?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/9d81ff1bea41</guid>
            <category><![CDATA[agentic-ai]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[business-strategy]]></category>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[ecommerce]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Tue, 13 Jan 2026 11:38:01 GMT</pubDate>
            <atom:updated>2026-01-13T11:38:01.155Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[AI needs faster memory systems. That just got invented]]></title>
            <link>https://chrisdare.medium.com/ai-needs-faster-memory-systems-that-just-got-invented-0d74af0dee6d?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/0d74af0dee6d</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[computing]]></category>
            <category><![CDATA[tsmc]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Sat, 03 May 2025 14:25:01 GMT</pubDate>
            <atom:updated>2025-05-04T07:51:05.969Z</atom:updated>
            <content:encoded><![CDATA[<h3>AI needs faster memory systems. They just got invented</h3><p>Hey world, it’s been a minute. I just discovered something worth sharing with you and wanted to drop this here. I think this will change the future of computing and adoption of AI as we know it. Enjoy.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*lalsXrdf2d1ZZo9b.jpg" /><figcaption>Image sourced from MIT Sloan Management Review</figcaption></figure><h3>The Need for Speed: Why Memory Matters</h3><p>Think about those moments when you’re waiting for a webpage to load, a file to download from the cloud, or for your AI-powered photo editor to work its magic. That little pause, that fraction of a second… it adds up, doesn’t it? Now, imagine that delay vanishing. And think about what that could change.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/350/0*Ul7TucNEm-4JjD-f.png" /><figcaption>Image sourced from <a href="https://www.computerhope.com/jargon/m/memory.htm">Computer Hope</a></figcaption></figure><p>Ever wondered why your phone can hold tons of photos but sometimes bogs down when you’re switching between apps? It’s pretty much because of 2 things: ROM and RAM. You might remember these terms from your computer class elementary school. Think of ROM as your phone’s permanent storage — all your files, safe and sound even when it’s off. Now, when you open an app, your phone pulls the necessary data into RAM — its super-fast, temporary workspace. More RAM means more apps running smoothly at once. Now the thing is RAM is lightning-fast but forgets everything when the power dies. This trade-off — speed versus permanent storage — has been the way it is for so long that it’s almost treated as an axiom in tech. For hungry AI, constantly needing to gulp down data, that speed bottleneck is a major hurdle because it matters how much high speed memory is available. In simple terms faster memory will enable faster AI, capable of quicker decisions — maybe even faster than us in critical situations like driving or surgery.</p><h3>Enter PoX Memory</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*QAYMWVHbLV7_bGyq.jpg" /><figcaption>POX memory researchers from Fudan University. Image sourced from 央视新闻 on Weibo</figcaption></figure><p>So what if we could bridge the gap? What if we could get memory that’s just as fast RAM or faster but can also hold onto data even when the power is off, like your regular storage? Well, a team of researchers in China at Fudan University might just have pulled off something incredible. They’ve developed a new type of memory which can read and write information at one-trillionth of a second. It’s just as fast as RAM and could even be up to 10 times faster. So this is a huge leap. They call it PoX (Phase-change Oxide) Flash Memory. However, this exciting breakthrough is still in its early stages, and the path to widespread use isn’t without its hurdles.</p><h3>Challenges and Realities</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*dYVmjMJfpRCxuWPu.jpeg" /><figcaption>Chip on a FAB manufacturing line. Image sourced from <a href="https://manufacturing-today.com/news/tsmc-commences-10-billion-semiconductor-plant-project-in-germany/">Manufacturing Today</a></figcaption></figure><p>This invention may still have a long way to go though. Turning a lab breakthrough into something you find in your next phone is a marathon, not a sprint. Scaling up production to meet global demand presents significant hurdles. For one thing, the materials used aren’t the same as what’s used to make RAM today. RAM today is mostly made of silicon, whereas this new memory is built with graphene. Silicon is the second most abundant element in the Earth’s crust, making it a relatively inexpensive raw material. Graphene, on the other hand, is derived from carbon sources like graphite, which is also abundant, but the process of creating high-quality graphene at scale has historically been more complex and costly. So making it cost-effective enough for mass adoption and integrating it smoothly into existing tech will take time and serious effort. Integration with existing chip architectures and software ecosystems will also require time and effort. That said there is tremendous commercial demand for this kind of technology. There will be bumps in the road, but those are challenges worth pursuing.</p><h3>Transformative Potential</h3><p>Think about the implications for AI. Some of the bottlenecks we’re currently facing in training complex models and deploying autonomous systems often come down to how quickly we can feed data to the processors. With memory this fast, AI could learn and respond at speeds previously thought impossible. When it comes to AI, speed goes a long way to build reliability. Models could be trained in a fraction of the time. Real-time AI analysis happening instantaneously, powering everything from more accurate medical diagnoses — where every millisecond counts. Or think truly intelligent autonomous vehicles that can process sensor data and react to hazards with superhuman reflexes.</p><h3>Wider Implications for Computing</h3><p>But the impact stretches beyond just AI. Consider cloud computing: faster memory in data centers could lead to lower latency, making everything from streaming services to collaborative work platforms feel snappier and more responsive. In scientific research, simulations that currently take days could potentially be completed in hours, accelerating the pace of discovery. Even in our everyday devices, like smartphones and laptops, this kind of speed could unlock new levels of performance and responsiveness, making multitasking seamless and demanding applications run effortlessly. Really exciting future.</p><h3>Final Thoughts</h3><p>The potential is so transformative that it’s worth paying close attention to the progress of PoX memory. If these initial speeds can be maintained and the manufacturing challenges overcome, we could be looking at a fundamental shift in the landscape of computing, with AI leading the charge into a future powered by high speed memory.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0d74af0dee6d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Running pytorch on apple silicon (M1) GPUs]]></title>
            <link>https://chrisdare.medium.com/running-pytorch-on-apple-silicon-m1-gpus-a8bb6f680b02?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/a8bb6f680b02</guid>
            <category><![CDATA[apple-metal]]></category>
            <category><![CDATA[deep-learning]]></category>
            <category><![CDATA[gpu-acceleration]]></category>
            <category><![CDATA[m1]]></category>
            <category><![CDATA[pytorch]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Sat, 21 May 2022 14:34:08 GMT</pubDate>
            <atom:updated>2022-05-22T08:43:37.508Z</atom:updated>
            <cc:license>https://creativecommons.org/licenses/by-sa/4.0/</cc:license>
            <content:encoded><![CDATA[<h3>Installing and running pytorch on M1 GPUs (Apple metal/MPS)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4Nxq9w1cnyIefm66-crylw.jpeg" /><figcaption>Image sourced from <a href="https://www.macrumors.com/2022/05/18/pytorch-gpu-accelerated-training-apple-silicon/">MacRumors</a></figcaption></figure><p>Hey everyone! In this article I’ll help you install pytorch for GPU acceleration on Apple’s M1 chips. Let’s crunch some tensors on Apple metal!</p><p>We’re in exciting times for the future of computing and AI. In 2020, when Apple released their new macbooks running on the M1 chips, one question on my mind with everything learned from Apple’s big reveal was <a href="https://github.com/pytorch/pytorch/issues/47702">whether pytorch could crunch tensors on those kickass GPUs and neural engines</a>. Fast forward to 2022 and the pytorch team <a href="https://pytorch.org/blog/introducing-accelerated-pytorch-training-on-mac/">announced a new version of pytorch that takes advantage of M1 GPUs</a>. Soumith and his team in collab with the Apple metal team have done an incredible job to make this possible. Kudos to them! While it’s early stage, it’s very exciting so far! Ideally, the instructions on the website should have been enough but I ran in to some roadblocks whilst setting up my mac to run pytorch on m1 GPUs. I am writing this article to provide guidance for anyone who might face similar issues. I hope you find this helpful.</p><h3>On installing python for m1…</h3><p>Tl-dr: Use conda</p><p>Installing python packages on the m1 macs itself is quite an extreme sport. I was at this for 2 days and a lot of tears when I set up my mac. I ended up resorting to installing and managing python packages on the the Rosseta’s x86 emulator for the terminal to run the intel version of homebrew and python versions and packages via pyenv. However, you can also build package wheels yourself if you’re up for it and have the time. My word of advice is to read lots of user experiences online before you start modifying your system🥲.</p><p>I use poetry to manage packages when building apps. When I am working on data science projects through, I prefer to use Anaconda. Conda’s great for data science and ML. You should be fine with either the GUI or installation script — but I opted to go with the GUI installation. As at the time of this writing, you will need to run pytorch on m1 GPUs via the preview (nightly) version of pytorch with the recommendation from pytorch’s website:</p><h3>Installing PyTorch</h3><p>Tl-dr: If you’re in a hurry and don’t want to read the rest of the article, then use the recommended pip wheel to install pytorch. It works!</p><pre>pip3 install --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu</pre><p>Verify your installation</p><pre>Python 3.10.4 (main, Mar 31 2022, 03:37:37) [Clang 12.0.0 ] on darwin</pre><pre>Type &quot;help&quot;, &quot;copyright&quot;, &quot;credits&quot; or &quot;license&quot; for more information.</pre><pre>&gt;&gt;&gt; import torch; torch.backends.mps.is_available()<br>True<br>&gt;&gt;&gt;</pre><p>Alright, now let’s see the details on the options available to us.</p><h3>Option 1: Install with conda</h3><p>At the moment, the pytorch recommends that you install pytorch, torchaudio and torchvision with conda.</p><pre>conda install pytorch torchvision torchaudio -c pytorch-nightly</pre><p>However, it’s ridden with errors (for now). For me in particular, I couldn’t find torchvision and torchaudio in the nightly channel for pytorch.</p><pre>Collecting package metadata (current_repodata.json): done</pre><pre>Solving environment: failed with initial frozen solve. Retrying with flexible solve.</pre><pre>Collecting package metadata (repodata.json): done</pre><pre>Solving environment: failed with initial frozen solve. Retrying with flexible solve.</pre><pre>PackagesNotFoundError: The following packages are not available from current channels:</pre><pre>- torchvision</pre><pre>- torchaudio</pre><pre>Current channels:</pre><pre>- <a href="https://conda.anaconda.org/pytorch-nightly/osx-arm64">https://conda.anaconda.org/pytorch-nightly/osx-arm64</a></pre><pre>- <a href="https://conda.anaconda.org/pytorch-nightly/noarch">https://conda.anaconda.org/pytorch-nightly/noarch</a></pre><pre>- <a href="https://repo.anaconda.com/pkgs/main/osx-arm64">https://repo.anaconda.com/pkgs/main/osx-arm64</a></pre><pre>- <a href="https://repo.anaconda.com/pkgs/main/noarch">https://repo.anaconda.com/pkgs/main/noarch</a></pre><pre>- <a href="https://repo.anaconda.com/pkgs/r/osx-arm64">https://repo.anaconda.com/pkgs/r/osx-arm64</a></pre><pre>- <a href="https://repo.anaconda.com/pkgs/r/noarch">https://repo.anaconda.com/pkgs/r/noarch</a></pre><pre>To search for alternate channels that may provide the conda package you&#39;re</pre><pre>looking for, navigate to</pre><pre><a href="https://anaconda.org/">https://anaconda.org</a></pre><pre>and use the search bar at the top of the page.</pre><p>I found though while there were no suitable packages found for torchaudio and torchvision, when I install pytorch only with conda install pytorch -c pytorch-nightly it works.I was also able to use the MPS backend. Right after that, you can install torchvision and torchaudio via their pip wheels</p><pre>pip3 install --pre torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu</pre><p>And then you’re good to go! Verify your installation by checking availability of the MPS backend in your python shell or jupyter notebook.</p><pre>Python 3.10.4 (main, Mar 31 2022, 03:37:37) [Clang 12.0.0 ] on darwin</pre><pre>Type &quot;help&quot;, &quot;copyright&quot;, &quot;credits&quot; or &quot;license&quot; for more information.</pre><pre>&gt;&gt;&gt; import torch; torch.backends.mps.is_available()<br>True<br>&gt;&gt;&gt;</pre><h3>Option 2: Install from pre-built wheels (Easiest and recommended)</h3><p>You can also install from pre-built wheels from the pytorch wheel archive. You’ll notice when you’re trying to install the stable version of PyTorch, the team actually recommends using the wheels to install the packages for M1 Macs</p><pre># MacOS Conda binaries are for x86_64 only, for M1 please use wheels</pre><p>What’s a wheel?</p><blockquote>A wheel is a type of <a href="https://packaging.python.org/glossary/#term-built-distribution"><strong>built distribution</strong></a>. In this case, <em>built</em> means that the wheel comes in a ready-to-install format and allows you to skip the build stage required with source distributions. — Brad Solomon, Real Python</blockquote><p>You can either install the wheels by <a href="https://download.pytorch.org/whl">downloading it from pytorch</a>/pip or building and installing a wheel from the source code itself. In this option, we we will download and install the pytorch wheel already built from the source code. You can do this by:</p><pre>pip3 install --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu</pre><p><em>The </em><em>--pre flag tells pip to install “Include pre-release and development versions. By default, pip only finds stable versions.”</em></p><p>Note: Be sure that you are executing pip for the virtual environment. I use pyenv to manage my python versions and found that pip3 was referring to my system pip. So I ran my command with pip instead. It depends on your setup</p><p>You’ll find that the MPS backend is also available when you check for it.</p><p>You learn more about python wheels through <a href="https://realpython.com/python-wheels/">Brad’s excellent primer</a> from RealPython.</p><h3>Option 3: Build pytorch (and torchaudio, torchvision) from source</h3><p>If you’re up for it, you can also build the packages from the source code. We will create our own wheels for pytorch, torchaudio and torchvision to install the nightly version of pytorch on M1 gpus.</p><p>Let’s build our binaries from source for installation. Here’s your guide curated from pytorch, torchaudio and torchvision repos. The documentation is yet to be updated for installation on MPS devices — so I had to make some modifications as you’ll see below:</p><p>Step 1: Create a conda environment</p><pre>conda env create --name pytorchm1<br>conda activate pytorchm1 </pre><p>Step 2: Configure installation dependencies</p><pre>### if you do not have a c++ compiler installed. You&#39;ll need it...<br>brew install gcc</pre><pre># install common dependencies<br>conda install astunparse numpy ninja pyyaml setuptools cmake cffi typing_extensions future six requests dataclasses</pre><pre># Add these packages if torch.distributed is needed<br>conda install pkg-config libuv</pre><p>Optional: Install mkl and mkl-include</p><p>I couldn’t install mkl and mkl-include but was still able to build my binaries for pytorch, torchvision and torchaudio. I ignored this error though since mkl and mkl-include are packages for optimizing math on intel CPUs and I am not running an intel CPU <a href="https://emojipedia.org/man-shrugging/">🤷‍♂️</a>.</p><p>But if you do need to check off the boxes, then:</p><pre>conda install mkl mkl-include</pre><p>…Or you could just build it from source :-)</p><p>If you’d like to learn more about mkl, check out the <a href="https://www.intel.com/content/www/us/en/develop/documentation/get-started-with-mkl-for-dpcpp/top.html">API docs</a>.</p><p>Step 3: Download sources for pytorch, torchaudio and torchvision</p><p>Once your conda env is activated, create a folder for your packages and then clone the repos into that folder</p><pre>mkdir packages &amp;&amp; cd packages<br>git clone --recursive <a href="https://github.com/pytorch/vision.git">https://github.com/pytorch/vision.git</a> torchvision<br>git clone --recursive <a href="https://github.com/pytorch/vision.git">https://github.com/pytorch/audio.git</a> torchaudio</pre><pre># get pytorch from the source <br>git clone --recursive <a href="https://github.com/pytorch/pytorch">https://github.com/pytorch/pytorch</a> pytorch</pre><p>You need to include submodules, which is why you should clone recursively.</p><p>Step 4: Install pytorch nightly</p><p>Installation instructions were sourced the <a href="https://github.com/pytorch/pytorch#from-source">pytorch readme</a>. However I made some edits to enable installation with Apple’s MPS. That’s just because I noticed that when you use the installation scripts from the pytorch readme, you’d get Apple’s MPS (<a href="https://developer.apple.com/documentation/metalperformanceshaders">Metal Performance Shaders</a>) deactivated by default. It can be changed with USE_MPS=1 USE_PYTORCH_METAL=1 (AlbanD mentions that</p><p>Let’s build PyTorch!</p><pre>cd pytorch<br># if you are updating an existing checkout<br>git submodule sync<br>git submodule update --init --recursive --jobs 0</pre><pre># checkout the nightly branch of pytorch<br>git checkout nightly # only if you want to use the preview build</pre><pre># install pytorch nightly<br>export CMAKE_PREFIX_PATH=${CONDA_PREFIX:-&quot;$(dirname $(which conda))/../&quot;}<br>MACOSX_DEPLOYMENT_TARGET=10.9 CC=clang CXX=clang++ USE_MPS=1 USE_PYTORCH_METAL=1 python setup.py install</pre><pre># cd back into your packages folder<br>cd ..</pre><p>You should have something like this at the tail end of your terminal outputs when it’s done:</p><pre>writing manifest file &#39;torch.egg-info/SOURCES.txt&#39;</pre><pre>Copying torch.egg-info to /Users/chrisdare/opt/anaconda3/envs/pytorchm1/lib/python3.9/site-packages/torch-1.13.0a0+git00c4d7f-py3.9.egg-info</pre><pre>running install_scripts</pre><pre>Installing convert-caffe2-to-onnx script to /Users/chrisdare/opt/anaconda3/envs/pytorchm1/bin</pre><pre>Installing convert-onnx-to-caffe2 script to /Users/chrisdare/opt/anaconda3/envs/pytorchm1/bin</pre><pre>Installing torchrun script to /Users/chrisdare/opt/anaconda3/envs/pytorchm1/bin</pre><p>Step 5: Install torchaudio</p><p>Installation instructions were sourced the t<a href="https://github.com/pytorch/audio#from-source">orchaudio readme</a></p><pre># install pending torchaudio dependencies<br>brew install pkg-config</pre><pre># move to your torch audio repo and install<br>cd torchaudio<br>CC=clang CXX=clang++ python setup.py install</pre><pre>cd ..</pre><p>You should have something like this at the tail end of your terminal outputs when it’s done:</p><pre>Installed /Users/chrisdare/opt/anaconda3/envs/pytorchm1/lib/python3.9/site-packages/torchaudio-0.12.0a0+6776299-py3.9-macosx-11.1-arm64.egg<br>Processing dependencies for torchaudio==0.12.0a0+6776299<br>Searching for torch==1.13.0a0+git00c4d7f<br>Best match: torch 1.13.0a0+git00c4d7f<br>Adding torch 1.13.0a0+git00c4d7f to easy-install.pth file</pre><pre>Installing convert-caffe2-to-onnx script to /Users/chrisdare/opt/anaconda3/envs/pytorchm1/bin</pre><pre>Installing convert-onnx-to-caffe2 script to /Users/chrisdare/opt/anaconda3/envs/pytorchm1/bin</pre><pre>Installing torchrun script to /Users/chrisdare/opt/anaconda3/envs/pytorchm1/bin</pre><pre>Using /Users/chrisdare/opt/anaconda3/envs/pytorchm1/lib/python3.9/site-packages<br>Searching for typing-extensions==4.1.1<br>Best match: typing-extensions 4.1.1<br>Adding typing-extensions 4.1.1 to easy-install.pth file</pre><pre>Using /Users/chrisdare/opt/anaconda3/envs/pytorchm1/lib/python3.9/site-packages<br>Finished processing dependencies for torchaudio==0.12.0a0+6776299</pre><p>Step 6: Install torchvision</p><p>The steps for installing torchvision are also similar</p><pre># install<br>conda install -c conda-forge ffmpeg<br>conda install pillow</pre><pre># install vision<br>cd torchvision<br>MACOSX_DEPLOYMENT_TARGET=10.9 CC=clang CXX=clang++ python setup.py install</pre><p>You should have something like this at the tail end of your terminal outputs when it’s done:</p><pre>Processing torchvision-0.13.0a0+b969cca-py3.9-macosx-11.1-arm64.egg</pre><pre>creating /Users/chrisdare/opt/anaconda3/envs/pytorchm1/lib/python3.9/site-packages/torchvision-0.13.0a0+b969cca-py3.9-macosx-11.1-arm64.egg</pre><pre>Extracting torchvision-0.13.0a0+b969cca-py3.9-macosx-11.1-arm64.egg to /Users/chrisdare/opt/anaconda3/envs/pytorchm1/lib/python3.9/site-packages</pre><pre>Adding torchvision 0.13.0a0+b969cca to easy-install.pth file</pre><pre>Installed /Users/chrisdare/opt/anaconda3/envs/pytorchm1/lib/python3.9/site-packages/torchvision-0.13.0a0+b969cca-py3.9-macosx-11.1-arm64.egg<br>Processing dependencies for torchvision==0.13.0a0+b969cca<br>Searching for torchvision==0.13.0a0+b969cca<br>Reading https://pypi.org/simple/torchvision/</pre><pre>/Users/chrisdare/opt/anaconda3/envs/pytorchm1/lib/python3.9/site-packages/pkg_resources/__init__.py:122: PkgResourcesDeprecationWarning:  is an invalid version and will not be supported in a future release<br>warnings.warn(<br>No local packages or working download links found for torchvision==0.13.0a0+b969cca</pre><pre>error: Could not find suitable distribution for Requirement.parse(&#39;torchvision==0.13.0a0+b969cca&#39;)</pre><p>The error at the bottom has to do with versioning and installing dependencies for torchvision. You may need to handle dependency installations manually. It should be easy with a conda or pip install.</p><h3>Let’s crunch some tensors!</h3><p>To check if the backend is available, use</p><pre>torch.backends.mps.is_available()</pre><p>Test your pytorch installation by running the <a href="https://pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html#sphx-glr-beginner-blitz-cifar10-tutorial-py">example image classifier code</a> from the pytorch website.</p><p>Be sure to use mps as the device for pytorch:</p><pre>DEVICE = torch.device(&quot;mps&quot; if torch.backends.mps.is_available() else &quot;cpu&quot;)<br>## and then move your model and data to the device before you train or eval. Have fun folks!</pre><p>And that’s it! You’re done installing pytorch to run on m1 GPUs. Have a good time crunching tensors!</p><h3>One more thing: Staying in the loop</h3><p>The next stable version of python is expected to have support for the MPS backend.<br>To learn more, and stay updated, visit the <a href="https://discuss.pytorch.org/c/metal-performance-shader/38">pytorch discuss forum for MPS</a>.</p><p>Shouts to <a href="https://discuss.pytorch.org/u/alband/summary">AlbanD</a>, you’re doing an awesome job!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a8bb6f680b02" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Face Recognition: Privacy and Bias Concerns]]></title>
            <link>https://medium.com/swlh/face-recognition-privacy-and-bias-concerns-c3faa7c54347?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/c3faa7c54347</guid>
            <category><![CDATA[privacy]]></category>
            <category><![CDATA[bias]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[face-recognition]]></category>
            <category><![CDATA[privacy-protection]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Sat, 05 Dec 2020 21:06:33 GMT</pubDate>
            <atom:updated>2022-08-27T17:00:41.898Z</atom:updated>
            <content:encoded><![CDATA[<h3>Face Recognition: Bias and Privacy Concerns</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pgCD1BZSqWAfUWmvL81dJg.jpeg" /><figcaption>The images in the header section were sourced from Unsplash</figcaption></figure><p><em>An audio version of this article is available on soundcloud. </em><a href="https://soundcloud.com/dexios1/face-recognition-privacy-and-bias-concerns">https://soundcloud.com/dexios1/face-recognition-privacy-and-bias-concerns</a>.</p><h3>Introduction</h3><p>Seminal improvements to face recognition technology in recent years have caused it to gain a seat at the table of biometric authentication and recognition innovations. Many countries have taken different approaches to the adoption and deployment of face recognition technologies. Consider opposing ends of the spectrum: Some countries (such as China) widely embrace its use while European countries under the GDPR are constrained by policy and legislation regarding its use. Other countries whose technology has fast outpaced their law, such as the USA, have quickly begun to develop additional policies and legislation regarding the use of facial recognition and data collection required to make it possible. [1,2] In this article, I highlight some concerns relating to bias and privacy in the deployment and use of face recognition technologies. Let’s fire away!</p><h3>Racial and gender bias in face recognition</h3><p>Research has shown that bias inherent in machine learning and AI datasets/models has the ability to reflect and cause harm in matters relating to race and gender [3]. This became a hot topic in natural language processing (also my favorite part of AI) but has now been proven in areas of computer vision — specifically, face recognition. Recent studies have also highlighted some high risks associated with the current SOTA face recognition technologies — especially with regards to gender misclassification. In a famous paper, Joy Buolamwini and Timnit Gebru demonstrated the bias in face recognition towards gender features that result in misclassification of individuals. Face recognition technologies such as Face++ as well as Microsoft’s Face API and IBM’s Watson Visual recognition application all yield significantly high classification errors of dark skinned women with rates as high as 36% [4]. Joy later shared the results of sample tests in her poem “AI, Ain’t I A Woman?” on YouTube. It was seen that face recognition technologies well respected by the public (at the time) gave high probability scores in their gender misclassifications of prominent and influential women in society. Images of famous personalities such as Oprah Winfrey, Michelle Obama, Serena Williams and Shirley Chisholm were all misclassified as men with accuracy ratings as high as 89% [5]. In use cases such as search or identification, this could likely place this category of individuals on the wrong/unjust side of the law or cause them to be denied access to a utility or service which they rightfully deserve to take advantage of.</p><h3>Privacy risks</h3><p>The law also needs to grow with the ubiquity of face recognition in order to meet the security and privacy threats that this technology faces. One such is deepfakes. Deepfakes use deep learning, a prominent form of artificial intelligence, to map the face of one person unto another person. Think of it like swapping faces — not only in photos, but videos as well. One of the most famous deepfake videos on the sane internet is the impersonation of Barack Obama by Jordan Peele…and oh boy, this version of Obama said some interesting things[6]. The prominence of deepfakes in 2019 also brought attention — as well as criticism; privacy and security concerns — to facial recognition technologies — especially in China where face recognition has been widely adopted for many important use cases in governance and business (e.g. surveillance and payment authorization) [7]. The Beijing News in particular, highlights the situation in China. They reveal that “artificial intelligence is not only a technical problem, but also a governance examination question”. They allude that personal privacy is growing scarce especially with the advent of face recognition and the emergence of deepfakes as seen in the Chinese app Zao (at the time of their writing). What happens when someone steals your face and uses it for nefarious purposes? The news agency concluded that legislation needs to catch up with artificial intelligence and companies currently need to observe ethical concerns in the design and use of AI models for face recognition[8]. (Now it’s important to note, that China is currently in the process of enacting their data protection and privacy law.)</p><h3>Final thoughts</h3><p>The fight to ensure a fair and balanced use of facial recognition technologies is not yet over. However there have been promising responses and results. We have since seen corporations respond to this call. Companies such as Alipay have added additional verification to determine whether an image being used to verify a payment is from a video, image or a fresh live scan of a person’s face[9]. That’s a great step to progress. Next up would be accurately detecting deepfakes. In the US, Amazon also published best practices for using its facial recognition API [10]. Following police brutality against people of color earlier this year (2020) and the accompanying protests, big tech companies such as Amazon, Microsoft and IBM have discontinued programs providing law enforcement agencies with access to their facial recognition technologies. While these decisions have been implemented in many flavors, it is still a step in the right direction. Many critics do, however, deliberate whether this will only be a 9-day wonder. [11]</p><p>Face recognition, just like many other technologies, promises a great way to ease our living and activities. However, until we develop the right laws and policies to view and manage this technology, it may cause more harm than good. There are many lessons we can learn from our experiences with good and bad cases of the use of this technology. Ultimately, it is of uttermost importance that the voice of the public is highly considered in the design and deployment of such technologies.</p><h3>Joy’s poem</h3><blockquote><em>“Obliged to you for hearing me, and now old Sojourner ain’t got nothing more to say.” — Sojourner Truth.</em></blockquote><p>I’ll conclude with Joy’s poem: “AI, ain’t I a woman?” Enjoy:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FQxuyfWoVV98%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DQxuyfWoVV98&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FQxuyfWoVV98%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/affda39ce5a9e7432c9801fd920ffc23/href">https://medium.com/media/affda39ce5a9e7432c9801fd920ffc23/href</a></iframe><h3>Other notes</h3><p>A lot has happened since I drafted this article in October 2020. A lot will happen after I publish it as well. I will use this section to provide additional notes which I believe are relevant and, as Jack Dorsey puts it, “point to a broader conversation” on this topic. I will also hyperlink references here instead of using IEEE style as I feel they are more appropriate for this exploratory section of the article.</p><h4>November 2020</h4><p>We had a talk concerning Bias in AI at CMU-Africa with the GIZ’s FAIR Forward and Stanford’s Renata Avila. Renata’s presentation was very informative. As aspiring researchers and policy makers, our students took pointers concerning designing and developing ‘fair machine learning models’. Amongst the many things she discussed, she shared information on efforts some researchers are making to mitigate bias in AI models. If you’re interested in learning more about it, read it <a href="https://controlling-bias.github.io/">here</a>.</p><p><strong>December 2020<br></strong>I was just about to publish this article when the bomb dropped: <strong>Timnit Gebru gets ‘fired’ from Google. </strong>$%#@!?!</p><p>Backstory: <em>Timnit Gebru is a world renowned AI researcher best known for her work in the bias and ethics of AI. I mention her paper concerning accuracy disparities in gender classification in this article[4]. Timnit and a few colleagues were asked by staff at Google to dissociate Google from a paper they co-authored for some reasons. Timnit gave them an ultimatum to drop it or she leaves. Google agreed to the latter.</em></p><p>The details of her departure from Google are out of the scope of my article and there’s also a lot we still don’t know surrounding these circumstances. Consequently, I will refrain from giving or linking details so as not to accidentally spread misinformation.</p><p>But here’s why I bring this up: Because of how important Timnit’s work is in the area of bias in AI, I believe this will draw a lot of attention to her papers…which is a good thing. Hopefully, we will still see some collaboration between Jeff Dean and Timnit in the future; <a href="https://twitter.com/JeffDean/status/1231336710161629184">Twitter convos, at least</a>. The paper she recently co-authored is titled <em>“On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?” </em>If you have a copy/link to the preprint, kindly leave it as a comment to this article. I look forward to reading it myself. <br>If you’d like to read Timnit’s other papers, <a href="https://paperswithcode.com/author/timnit-gebru">check them out at PapersWithCode</a>.</p><p>What a way to enter Neurips this week though!</p><p><strong>New note to data scientists researchers, ML engineers/enthusiasts:</strong></p><h4><strong>NeurIPS 2020!!</strong></h4><p>At the time I’m writing this NeurIPS 2020 (nips) has just kicked off!!! Super exciting! (Unfortunately for me, I have to juggle between nips and CMU homeworks :-\ ). We’re 40 minutes into nips and both ongoing presentations are hitting the nail on the head when it comes to fairness in AI. Speakers from Netflix and Scikit-learn are discussing ways to mitigate bias in reinforcement learning (Netflix) as well as traditional machine learning (sci-kit learn). I would try to provide updates on nips with respect to bias and privacy sometime later. FYI, p(doing_this)=0.4</p><h3>References</h3><ol><li>BBC News, “Facial recognition: School ID checks lead to GDPR fine”, 2019. [Online] Available: ​ <a href="https://www.bbc.com/news/technology-49489154">https://www.bbc.com/news/technology-49489154</a>​ [Accessed: 13-Oct-2020]</li><li>116th Congress (2019–2020), “S.847–116th Congress (2019–2020): Commercial Facial Recognition Privacy Act of 2019 | Congress.gov | Library of Congress”, 2019. [Online] Available: ​ <a href="https://www.congress.gov/bill/116th-congress/senate-bill/847">https://www.congress.gov/bill/116th-congress/senate-bill/847</a>​ . [Accessed: 14-Oct-2020]</li><li>T. Bolukbasi, Kai Wei Chang, J. Y. Zou, V. Saligrama and A. Kalai, “Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings” in Computing Research Repository (CoRR), 2016.</li><li>J. Buolamwini and T. Gebru, “Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification” in Proceedings of Machine Learning Research no. 81, pp. 1–15, 2018.</li><li>J. Buolamwini, “AI, Ain’t I A Woman? — Joy Buolamwini — YouTube”, 2018. [Online] Available: ​ <a href="https://www.youtube.com/watch?v=QxuyfWoVV98">https://www.youtube.com/watch?v=QxuyfWoVV98</a>​ [Accessed 12-Oct-2020]</li><li>BuzzFeedVideo, “You Won’t Believe What Obama Says In This Video! 😉” on YouTube, 2018. [Online] Available: <a href="https://www.youtube.com/watch?v=cQ54GDm1eL0">https://www.youtube.com/watch?v=cQ54GDm1eL0</a></li><li>G. Shao and E. Cheng, “The Chinese face-swapping app that went viral is taking the danger of ‘deepfake’ to the masses”, 2019. [Online] Available:<br><a href="https://www.cnbc.com/2019/09/04/chinese-face-swapping-app-zao-takes-dangers-of-deepfake-to-the-masses.html">https://www.cnbc.com/2019/09/04/chinese-face-swapping-app-zao-takes-dangers-of-deepfake-to-the-masses.html</a> [Accessed 14-Oct-2020]</li><li>The Beijing News, “AI changes faces to cause controversy, user privacy can not be violated”, 2019. [Online] Available:<br><a href="http://epaper.bjnews.com.cn/html/2019-09/01/content_763972.htm?div=-1">http://epaper.bjnews.com.cn/html/2019-09/01/content_763972.htm?div=-1</a>​ . [Accessed: 14-Oct-2020]</li><li>G. Shao and E. Cheng, “Growing backlash in China against A.I. and facial recognition” 2019. [Online] Available:<br><a href="https://www.cnbc.com/2019/09/06/ai-worries-about-the-dangers-of-facial-recognition-growing-in-china.html">https://www.cnbc.com/2019/09/06/ai-worries-about-the-dangers-of-facial-recognition-growing-in-china.html</a> [Accessed 14-Oct-2020]</li><li>Amazon Web Services, Inc, “Use cases that involve public safety”. [Online] Available: <a href="https://docs.aws.amazon.com/rekognition/latest/dg/considerations-public-safety-use-cases.html%E2%80%8B">https://docs.aws.amazon.com/rekognition/latest/dg/considerations-public-safety-use-cases.html​</a> . [Accessed: 12-Oct-2020]</li><li>K. Hao, “The two-year fight to stop Amazon from selling face recognition to the police”, 2020. [Online] Available: <a href="https://www.technologyreview.com/2020/06/12/1003482/amazon-stopped-selling-police-face-recognition-fight/%E2%80%8B">https://www.technologyreview.com/2020/06/12/1003482/amazon-stopped-selling-police-face-recognition-fight/​</a> [Accessed: 12-Oct-2020]</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c3faa7c54347" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/face-recognition-privacy-and-bias-concerns-c3faa7c54347">Face Recognition: Privacy and Bias Concerns</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Qualities of Leaders We Need]]></title>
            <link>https://chrisdare.medium.com/the-qualities-of-leaders-we-need-28796703f469?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/28796703f469</guid>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[carnegie-mellon]]></category>
            <category><![CDATA[elections]]></category>
            <category><![CDATA[education]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Fri, 22 Nov 2019 22:44:17 GMT</pubDate>
            <atom:updated>2019-11-23T13:08:02.859Z</atom:updated>
            <content:encoded><![CDATA[<p>This is an article I wrote for the classes of 2020 and 2021 of Carnegie Mellon University in Africa. I decided to put this here on medium because I believe this piece of writing is one that anyone who has given the power to elect a leader need to consider in their decision-making. Let’s begin.</p><p>Hello Tartans,</p><p>We’re finally at that time of the year where it’s time to appoint student representatives to be at the helm of our affairs. As promised, I am sharing with you wisdom concerning the qualities of leaders we need. I’d been rethinking what I initially planned to share with you. Because, quite frankly, I felt some of these are things you have at least heard before. Also, I didn’t want to bombard you with large blobs of text. So I deleted the text I wrote here on Monday and rather asked myself, “If I had to mention one quality, what would it be?” I got my answer…but even for me, as the author of this text, it was not quite what I expected for two reasons. Let me mention what that quality is first, and then we can get to that:</p><blockquote><strong>The leaders we need ensure that they carry out the vision for progress</strong></blockquote><p>Here’s why it was a surprise to me: First, this is not really a quality but more of a characteristic. Secondly, I realized that it was less about the running candidates <strong>and more about you.</strong></p><p>There’s a statement President Kagame made last month. I’d like to quote it because it’s very relevant to this characteristic/quality. He said:</p><blockquote>“You don’t have to look to government and say government is not doing this, government is not good at doing this. But government is actually you,”</blockquote><p>Government exist at various levels and in different organizations. Here at CMU, and in organizations you will build/work with, it will be no different. <strong>Government is you</strong>; you are a leader too and you need to carry out the vision for progress. When we finally choose from our fellow students to lead us by service in our guild, <strong>don’t leave them to do the work alone.</strong> That means supporting them, taking up responsibilities, challenging them in a<strong> wise and constructive manner</strong> when you think they are taking the wrong path…and more.</p><p>The moment you put someone in power, <strong>you have done them a favor</strong> (and not vice versa). In our context, you’ve given them something honorable to slap on a resume/CV, an opportunity for their self-improvement and/or some other things that will benefit <strong>them</strong>. Everything else that happens after is supposed to be about <strong>you</strong>. But don’t fold your arms to be just an observer or a critic. If that happens, we will lose productivity and never know we did! Whatever happens after you put people in power, I argue that you should see it as a responsibility to help them because that is helping you. <strong>Whether or not you will be recognized for your efforts, it will surely build you up and prepare you for something greater that’s ahead.</strong></p><p>Shalom.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=28796703f469" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The limits and future of AI]]></title>
            <link>https://medium.com/sigma-ai/the-limits-and-future-of-ai-bc7d64f189f3?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/bc7d64f189f3</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[artificial-neural-network]]></category>
            <category><![CDATA[research]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Thu, 24 Oct 2019 11:21:02 GMT</pubDate>
            <atom:updated>2019-10-26T19:36:57.648Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*pef1zv49BQHx9Q8Z.jpeg" /><figcaption>Illustration of a neural network. Source: <a href="https://tech.kartenmacherei.de/classifying-events-using-a-neural-network-488acb50de87">https://tech.kartenmacherei.de/classifying-events-using-a-neural-network-488acb50de87</a></figcaption></figure><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplay.ht%2Fembed%2F%3Farticle_url%3Dhttps%3A%2F%2Fmedium.com%2F_p%2Fthe-limits-and-future-of-ai-bc7d64f189f3&amp;url=https%3A%2F%2Fplay.ht%2Farticles%2Fbc7d64f189f3&amp;image=https%3A%2F%2F2%2ADL5_5fBu0GjX3oVtb5nHwA.jpeg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=play" width="700" height="185" frameborder="0" scrolling="no"><a href="https://medium.com/media/02695ea58ff7d1d805d81a553d69d304/href">https://medium.com/media/02695ea58ff7d1d805d81a553d69d304/href</a></iframe><blockquote>“Any sufficiently advanced algorithm is indistiguishable from intelligence” — Yours truly</blockquote><p>Many would be right to say that this article deserves a place in an opinion column of a publication. Nonetheless, it’s all based on known facts and insights from some of the best professors in the world. It’s also a complex subject and so I’ve tried as much as possible to describe what I refer to from a high level of abstraction.</p><p>Let’s begin.</p><p>These days we see or hear in the news, a lot of concerns, shrieks, and complaints like the following:</p><h3>“AI is taking over the world! AI is dangerous! Machines are becoming more intelligent than us…” (sighs)</h3><p>Hold your horses. We’re not there yet. In this article, I will juxtapose human and machine intelligence to reveal the gaps between these two. I will also predict what will be required to make a huge leap to the next stage of the evolution of AI. Consider this a primer, not a deep dive into the subject matter.</p><p>What comes to your mind when you think about AI? A robot that can converse with you, cook your lunch? Sure, that counts. What also counts is Gmail; it’s able to recommend what you should type next as you compose your email — or let’s also consider tagging on Facebook; you upload a photo of you and some friends and Facebook automatically tells you that your friend John is also in the photo. (There are several examples.) Interesting right? Well, maybe that’s AI — and maybe it’s not. The thing is, terms like artificial intelligence (AI), machine learning (ML), deep learning (DL), and even deep neural networks (DNNs) have become buzzwords today. Moreover, their definitions keep changing over time. The public and scientific opinion of AI from the 1950s is different from what it is today. Once AI meant “search”, then it changed to expert systems…and now it’s centered around neural networks (and deep learning).</p><p>So can we establish a sound definition of AI? Probably. Scientific peers will agree that <a href="https://www.amazon.com/dp/9332543518/">we can define an AI entity as an agent that is capable of reasoning, perception, and acting on its environment</a>. Then we can also throw in the fact that’s not a biological agent. Otherwise then a mimosa plant would also qualify as AI. You can also think of AI as anything that is intelligent but not biological — for simplicity.</p><p>And then comes the first fundamental issue. Take a look at these words below:</p><blockquote>Artificial Intelligence<br>Machine Intelligence<br>Computational Intelligence</blockquote><p>We can even go on to reference terms such as:</p><blockquote>Machine Learning<br>Knowledge-based systems</blockquote><p>I think that’s interesting. You know why? Because intelligence, learning, and knowledge are all human attributes! They’re actually not meant to describe machines. A machine that <a href="https://www.youtube.com/watch?v=r5Ju2QuYpwQ">functions exactly like a mimosa plant</a> will today be considered as intelligent by many. But do we really identify an actual mimosa plant as intelligent?</p><p>It’s hard to define intelligence in the context of machines. But that’s what we’re trying to do after all, right? We want to give machines all/some qualities of humans. So, has this been successful? Let’s find out.</p><p>In order to give machines the level of intelligence humans possess, we first need to understand how human intelligence or <a href="http://vernon.eu/ACS.htm">cognition</a> works. What comes into your mind when you think about human intelligence? What’s one important component that comes to mind? The brain? Yes, the brain. Well apparently, it turns out that we don’t know a lot about the brain and how it really works. It seems that the more we try to study the brain, the more we realize that we know very little about how it works. (Hopefully, that changes in the coming years.)</p><p><strong>Side note:</strong> For the purpose of comparison and generalization, I may use the “brain” or “human intelligence” or “biological neural networks (BNNs)” interchangeably. I may do the same with the terms “AI” and one of its subfields — <a href="https://www.indiabookstore.net/isbn/9788120312531">artificial neural networks</a> (ANNs). <em>An artificial neural network is simply a model inspired by the structure and function of a biological system in order to perform certain tasks that require intelligence.</em> Not sure what a model means? In the context of tech and AI, It’s basically a representation of a system — often it’s mathematical. (We normally use math in such representations because unlike other languages, math is unambiguous.) Now, let’s get back to where we left off.</p><p>So I was telling you about the brain. The subject of the brain is very important because in its architecture lies a fundamental difference between biological neural networks (humans in this case) and artificial neural networks (machines).</p><p>Computers are based on a type of architecture called the Von Neumann or Princeton model. Here’s the diagrammatic representation:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*W1r01jqMWu-kENd8.png" /><figcaption>Source: Wikipedia. It’s a 74-year-old architecture; Wikipedia was sure to get it right and they had a great diagram ;-)</figcaption></figure><p>The interesting thing is that this model was invented as far back as 1945 and computers today still have this architecture. They basically load a sequence of instructions from memory and execute them in the CPU. That hasn’t changed since 1945; we’ve only made computers more sophisticated in a manner that allows them to fetch, decode and execute programs faster. Even many <a href="https://www.researchgate.net/search?q=Quantum%20von%20Neumann%20Architecture">emerging quantum computers are being built this way</a>. The point is the architecture is still the same.</p><p>With the brain, on the other hand, …well we don’t even know how the brain really works?!? Between the brain and machines/computers, what we do know however is that memory works fundamentally differently. When I want to recall something like a friend’s full name, I just need to think about a part of what I’m looking for, a first name perhaps, and boom — I’ll remember. It’s not the same with computers. You have to reference an object in computer memory by an address. Imagine you wanted to recall what you ate last Sunday and you had to think of its address code — something like 114562 — in order to remember it. :-(</p><p>Well, that’s how computers work right now. This leads us to a limitation of the Von Neumann model:</p><blockquote>You can’t do processing in memory</blockquote><p>So far, we know that what makes biological neural networks advanced is that neurons are both memory and processing — unlike the Von Neumann machines we use today. Think of the brain like a computer for whose CPU, RAM and all other kinds of storage are the same component. We can call that a <a href="https://ieeexplore.ieee.org/document/7551380">processing-in-memory (PiM) model</a>. When genuine PiM models/architectures are successfully implemented, they will change the way the computer industry operates. Chip, phones, desktop, and laptop manufacturers…even AI companies &amp; researchers, software developers and other professionals in the industry will need to adapt — or die. A lot of research has been put into realizing a PiM architecture but it’s hard to tell when a breakthrough will occur.</p><p>The thing is the latest advance in artificial intelligence is pretty much artificial neural networks. However, using today’s machines to implement the computational model of a biological neural network leaves you with parallel distributed processing. And one of the reasons for this is that fundamental difference in the memory systems of biological architectures and the Von Neumann architectures.</p><p>For us to build a true PiM computer that can cause an AI evolution, we may need to deeply answer questions like: “<em>Why does the brain work the way it works?” </em>and<em> “Why is the brain built the way it is?” </em>These are important to understand because one needs to be sure that an imitation of the brain — in modeling such an architecture for AI — will be contextually relevant. It’s something I learned from <a href="https://www.researchgate.net/publication/235626691_The_vision_of_David_Marr">Marr’s level of abstraction</a>. It helps you know how and what you should be really modeling. Take, for example, the invention of aircraft. For centuries people had attempted to build aircraft by trying to copy birds. It looked easy; all you had to do was glue some feathers and flap your wings like an eagle — until you keep falling off the cliff. The problem was that people at the time where observing birds from the wrong level of abstraction. It turns out that a bird’s wings conform to the laws of fluid dynamics. What enabled the first airplane to fly was that it followed the same principles. However, rather than flapping wings, planes are propelled by engines. We took the good parts and ignored inefficient features that couldn’t take us to other domains like outer space. Now we’re trying back at this again, this time to model AI and computers after the human brain. Thus, it’s important to understand that observing the brain from the neural network level of abstraction is not enough. If we also unlock the secrets of the brain at the memory level of abstraction, perhaps we can model computers and neural networks in a manner that is fundamentally the same, different implementation yet realizing an extremely reasonable form of human intelligence. <a href="https://www.ibm.com/blogs/cloud-archive/2015/11/future-of-cognitive-computing/">IBM seems to understand this all too well</a>. They also have a <a href="https://www.research.ibm.com/software/IBMResearch/multimedia/Computing_Cognition_WhitePaper.pdf">paper</a> on this.</p><p>Probably the last thing I will mention is that you need to understand that the brain doesn’t exist on its own — it’s part of a body. As a result, one can argue that merely copying how the brain works without understanding how the rest of the body functions and how it’s connected to the brain can lead to a colossal fundamental failure. These are concerns researchers are addressing today in order to ameliorate the challenges in the evolution of artificial intelligence.</p><p>One final distinction about AI today I need to mention is that while humans require little data to do numerous tasks, AI tools of today require massive amounts of data in order to perform domain-specific tasks.</p><p>This is where I’ll stop. There are far more things to say but this has been a good enough primer to address the concerns many people have about AI. The heights attained by AI is truly remarkable. However, there’s a huge gap between human intelligence and machine intelligence. AI is currently just a tool that’s smart enough to execute a domain-specific task. We’re fortunate to still be in the early stages of its development and as we make breakthroughs it’s important that we model AI in a manner that augments our capabilities rather than shooting us in the foot.</p><blockquote>“The only way of discovering the limits of the possible is to venture a little way past them into the impossible.” — Arthur C. Clarke</blockquote><p>I&#39;m on twitter. If you loved this article, <a href="https://twitter.com/chrisdare_">follow me for updates on new releases</a>.</p><p>Cheers.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bc7d64f189f3" width="1" height="1" alt=""><hr><p><a href="https://medium.com/sigma-ai/the-limits-and-future-of-ai-bc7d64f189f3">The limits and future of AI</a> was originally published in <a href="https://medium.com/sigma-ai">Sigma AI</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Transforming Africa’s Commodity Market with Digital Technology]]></title>
            <link>https://chrisdare.medium.com/transforming-africas-commodity-market-with-digital-technology-d5eaeae7fb30?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/d5eaeae7fb30</guid>
            <category><![CDATA[africa]]></category>
            <category><![CDATA[cmu-africa]]></category>
            <category><![CDATA[commodities]]></category>
            <category><![CDATA[digital-transformation]]></category>
            <category><![CDATA[agriculture]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Fri, 20 Sep 2019 09:33:41 GMT</pubDate>
            <atom:updated>2019-10-14T08:52:24.615Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*1Z8y-hMIumbGXY9T" /><figcaption>From left to right: Wuyeh Jobe, Bonaventure Twagirimana, Yours truly and Rahma Ahmed</figcaption></figure><p>We had a fantastic guest lecture yesterday at <a href="https://www.linkedin.com/school/167009/?legacySchoolId=167009">Carnegie Mellon University Africa</a> with Bonaventure Twagirimana, head of ICT and Market Data at the East African Exchange (EAX). This is part of a course offered at CMU-Africa: Strategic Use of Digital Information in Enterprises. It’s taught by <a href="https://www.linkedin.com/in/mbezy/">Michel Bezy</a>, Distinguished Professor and former Program Director at IBM.</p><p>Bonaventure (CMU Africa alumnus) gave insights into how the exchange is transforming the commodity market in East Africa through the use of digital technology. Some examples are the use of sensors to detect and maintain high-quality grain, cloud computing and highly effective NASDAQ trading systems.</p><p>The results of their work so far? High-quality grains, connecting farmers to capital markets, and the securitization of grains: spots, options, forwards and futures that can be traded on international markets! With over $3mn in revenue, 5 years in operation, and being a leader with only 2% of the market share, I can only conclude that the future is very bright for the exchange. The exchange plans to expand into managing other crops like coffee and tea as well as increase adoption rates by farmers in the country.</p><p>For farmers, this is good news. Many farmers in Africa often sell their products too quickly in order to ensure that they have the cash and capital for the next planting season as well as living expenses. As a result, products are often undervalued since they are also sold in times of influx supply. But now, with the help of the Exchange, these farmers can securitize their harvests and ultimately maximize their returns. Let me explain further: With the help of the Exchange, these farmers can store their grain in a warehouse managed by the EAX. Then they receive a virtual receipt that represents their physical product. Farmers can then use that receipt to access capital from financial institutions like banks since the crop is guaranteed as an asset.<br><em>Here’s what that means: The receipts/contracts can be sold based on the present/future price of the crop. They can be also be used to access loans because the product in the warehouse managed by EAX is now considered as an asset.</em></p><p>For investors and financial institutions in East Africa and Rwanda, this is even juicier when you compare with other numbers: Lending rates in Rwanda average 19%pa, Fixed deposits return 8–10%…whilst trading commodities can return an average of 20–30%pa for initial resellers.</p><p>(If you are new to the concept of commodities trading, I must admit that my explanation may be a little hard to grasp. Due to the scope of my post and the limited time I have to write these days, I won’t go into the details. That said, I may reword it later or illustrate it with a diagram. In the meantime then, take a look at <a href="https://www.investopedia.com/investing/commodities-trading-overview/">this link</a> to learn more or read chapter 2 of Ray Dalio’s book “Principles. If you felt my explanation was also fine, do let me know in the comments so I save my time. Thanks.)</p><p>I had an “after-party” discussion with Bonaventure and my stellar mates <a href="https://www.linkedin.com/in/ACoAABrs8JsBzrPfbLLjnJ-uLP_iUVGjPGeWzZE/">Wuyeh Jobe</a> and <a href="https://www.linkedin.com/in/ACoAAAj9zfkBmzrSu7chNOleI5UnKxx48o4GIY0/">Rahma Ahmed Samaroon</a> and he revealed the key role governments (and other investors) have to play in supporting farmers and exchanges all over Africa. We also talked about the applications of machine learning to predict the price of grains and more.</p><p>Similar work is also taking place in West Africa with the launch of the GCX in Ghana by a founding member of the East African Exchange. These are exciting times for markets in Africa.</p><p>Here’s a worthy note by a friend and <a href="https://seo-africa.org/">SEO Africa</a> ’17 Alum, <a href="https://www.linkedin.com/in/desmonddarlington/">Desmond Darlington</a>: <em>“Ethiopia started this though. So if there’s a sequel, acknowledge their pioneering spirit”</em></p><p>Ethiopia, you are fully acknowledged.💯</p><p>There’s also one name that runs through the success and operations of these commodity exchanges in East Africa, Ethiopia, and Ghana: <strong>Dr. Kadri Alfah.</strong></p><p>This was a short article; I have little time to write for the web these days. Leave a few claps to let me know if you enjoyed the read. I’m deliberately abstracting the contents of the article so it’s brief and since it’s my first time doing this, I’d like to get some feedback on it via comments or on WhatsApp/Skype. Thanks in advance.</p><p><strong>If you’d like to learn more about Carnegie Mellon University in Africa and our new campus, click here: </strong><a href="https://www.cmu.edu/news/stories/archives/2019/august/cmu-africa.html"><strong>https://www.cmu.edu/news/stories/archives/2019/augus</strong>t/cmu-africa.html</a>. Applications are also open for Ms. in Information Technology as well as Electrical Engineering and Computer Science. Visit <a href="https://cmu.edu/africa">https://cmu.edu/africa</a> to start your application today.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d5eaeae7fb30" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Hello Kigali (A Beginner’s Guide)]]></title>
            <link>https://chrisdare.medium.com/hello-kigali-a-beginners-guide-cf66f94cabf?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/cf66f94cabf</guid>
            <category><![CDATA[cmu-africa]]></category>
            <category><![CDATA[rwanda]]></category>
            <category><![CDATA[kigali]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Sun, 04 Aug 2019 19:12:59 GMT</pubDate>
            <atom:updated>2019-08-05T09:26:08.522Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RHhVkA04iwvCvhv8-z729w@2x.jpeg" /></figure><p>Bite!👋👋👋👋 (It translates from Kinyarwandan to “hello” in English)</p><p>It’s been a while I wrote something so I thought to break the long overdue silence by sharing a few experiences from my visit to Kigali in Rwanda. If you’re from West Africa (especially) and looking to visit Rwanda, this should be helpful. Bear in mind, however, that it’s not extensive. I’ve just been here for 2 days. But I’m publishing this story now because I’m not sure if I’ll write about it after a week or two passes. Got a tight schedule coming up. Also, this is just limited to Kigali because I haven’t been to any of the country’s other provinces yet.</p><p><strong>Ok people, let’s roll!!!</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FIkg97EyA-HlL5OtQfVezw@2x.jpeg" /><figcaption>Closest thing I’ve seen (with my eyes) to the aurora so I had to grab a shot.</figcaption></figure><p>I touched down in Kigali after a 6/7-hour flight from Accra. It was night time and I couldn’t see much but the streets gave me good vibes: they’re clean and lit (confirmed in daylight). I checked in to my hotel a few minutes later and went to have dinner.</p><h3>So, about dinner. . .</h3><p>Man, it’s tasteless😂😂.</p><p>“Chicken and chips,” I said to myself, “What could go wrong?”</p><p>Well, apparently everything🤣 🤣. I couldn’t taste a thing…lol.</p><p>Taste is everything in Ghana so the food here is off for me. But I have to get used to it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/0*We1BxRT_yPC-Zsox.jpg" /><figcaption>Image from Envato.</figcaption></figure><p>You’re gonna need these two guys up there especially if you’re West African. Everyone says I’ll get used to it — but I’m not sure yet. Food without the standard seasoning I know has been a bummer but I’m trying to improvise. Fried eggs taste the same though — after enforcing the cook to add salt. Nonetheless, even the most common of foods here in Rwanda is very healthy. Gotta say…I’m impressed, Rwandans.</p><p>There’s a lot of potato chips here too. Common and cheap. Like fried yam in my motherland, Ghana!!!!</p><h3>Weather?</h3><p>Is it cold or hot? Ummm it depends on where you’re coming from. So for me, it was a little cold for me at the start. Summer’s ending and we’re about to enter the dry season. Temperatures got to 15C sometime today but for the most part, it’s gonna be somewhere close to 21C.</p><p><em>(Now I wrote that yesterday — but it was pretty darn hot the day after. Can’t tell for sure yet. So what I’m gonna do is that I’ll just publish this story and update later on in the future. But then I can tell you that Kigali’s about to enter the dry season. So it’ll be pretty cold — for me — in the evenings.)</em></p><p>And…I haven’t seen much to shed more light so let’s keep it at that.</p><h3>Cost of Living?</h3><p>Since most people can relate to dollars, I’ll quote the prices in US Dollars and Rwandan Francs. You also have to note that Inflation or deflation, as well as interest rates, can cause these figures to change over time. So take note.)</p><p>So a bottle of water is pretty much 300RWF which is a little under 2GHS and 0.40USD. You can grab a 330ml bottle of Fanta or coke for 500RWF or 0.50USD.</p><p>Food? Uhh you pretty much have to choose between going to a restaurant or cooking. Unlike Accra or Lagos, there’s no waakye or suya seller by the roadside. Taxation is pretty tight here and you need to set up a restaurant if you wanna sell food. They’re gonna check your papers and assert that you’re paying tax and all of that stuff. A little over 10,000 RWF or 11 dollars a day can get you breakfast, lunch, and supper. Thinking about it now, I just realized I haven’t had lunch since I got here…interesting🤔. Breakfast has been awesome and I’m learning to stop complaining to myself about dinner. (I’m addicted to seasoning, charle.)</p><p>Transportation? If you’re just here to visit, use “on-demand” transport or get a tap-and-pay bus card. Your charge on the card starts at 300WF or 0.40USD. There isn’t Uber, Taxify or Yango here. Rather you can use this app called Move by VW or <a href="https://ktpress.rw/2018/09/yego-cab-a-new-low-cost-taxi-service-launched-in-kigali/">Yegocabs</a>. Now I’m not sure about Yegocabs’ pricing but Move will cost you about 400RWF or 0.50USD per km. Bikes are also very common in Kigali so you can get a rider to take you anywhere you wanna go. Now we know them as Okadas in Accra and Lagos. (Also here, TVS is the most common brand of bikes.) You can also get a bicycle rider to move you around if you wish — but I don’t trust its reflex capacity in moving traffic. Not sure about the rate either but it’s cheaper.)</p><p>Motor prices start at 300RWF or 0.40USD and if you’re going to someplace in Kigali you shouldn’t have to pay more than 1700RWF or 2 USD. Apparently, if you learn Kinyarwanda, the local dialect, you can get or negotiate better quotes with the riders. Otherwise, they’ll just slap mad prices on you.</p><p><em>About taxis…well the fares are simple but not that simple: </em><br>The first kilometer you travel will cost you 1500RWF or 1.7USD. <br>An additional 29km will cost you 700RWF OR 0.80USD per km. <br>Any kilometer after that costs you 500RWF or 0.60USD.</p><p><em>S/O to my CMU class for helping me put the transport rates together!</em></p><p>Accommodation? 80 dollars a night can get you a decent hotel (with breakfast, a pool, sauna, and a gym). 400–500 dollars a month can get you a 4 bedroom house to rent. Airbnb will be your best bet if you’re traveling alone and have a decent budget.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gBJ6xge7pGZT3Ae0pJHXmw@2x.jpeg" /><figcaption>4 bedroom house in Kigali costing somewhere about 450 USD</figcaption></figure><h3>Other realizations</h3><p>Energy seems like a big thing here in Kigali. I noticed a number of energy companies in Kigali’s economic zone. The use of Solar energy is ‘pretty’ common here too.</p><p>I really love the way Rwandans patronize their own products first and personalize their adoption of foreign technology. The government also runs a number of companies that work well. Not sure about their profitability but I think the numbers are positive.</p><p>There’s still a lot of land to build for business or residential purposes in Kigali.</p><p>Also, Rwandans have a really smart way of naming number plates. I think it pretty intelligent. (Number bases alert!) The last place for the car number is an alphabet rather than a number. So technically, you can say that the numbering is in base 26. That makes you accommodate a large set of numbers.</p><p>Street naming is so darn easy to remember.</p><h3>Advice for aspiring visitors from West Africa?</h3><p>Just pack seasoning and you’ll be fine😂</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/516/1*SfRV7hQbq0hFoIoHU0KztA@2x.jpeg" /><figcaption>Till next time!</figcaption></figure><p>Hey, by the way, can you help me retweet this? It’s an irritating problem I have which I think Elon Musk’s Neuralink or a competitor (I guess) can solve. Pretty sure it’s been thought of before but I just got the idea yesterday. I think we can use a brain to machine interface technology to help us remember people’s names from our first encounters with them: <a href="https://twitter.com/chrisdare_/status/1158082111737356288">https://twitter.com/chrisdare_/status/1158082111737356288</a></p><p>Retweet it…what could go wrong? You’ve read this story for free…now pay me with a retweet.🤷🤷 Thanks in advance! :-D</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cf66f94cabf" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Theresa May Resigns. Here’s why — ish.]]></title>
            <link>https://chrisdare.medium.com/theresa-may-resigns-heres-why-ish-23d5b1a2c518?source=rss-e763f8f5a8f1------2</link>
            <guid isPermaLink="false">https://medium.com/p/23d5b1a2c518</guid>
            <category><![CDATA[british-pound]]></category>
            <category><![CDATA[brexit]]></category>
            <dc:creator><![CDATA[Chris Dare]]></dc:creator>
            <pubDate>Fri, 24 May 2019 16:43:47 GMT</pubDate>
            <atom:updated>2019-05-26T06:47:12.015Z</atom:updated>
            <content:encoded><![CDATA[<h3>Theresa May Resigns as UK Prime Minister. Here’s why — ish.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/360/1*eS6-8tWCTsOBFdDR40MR_g@2x.jpeg" /></figure><blockquote>“Ever since I first stepped through the door behind me as Prime minister I have striven to make the United Kingdom a country that works not just for a privileged few but for everyone — and to honor the results of the EU referendum…Our politics may be under strain but there is so much that is good about this country; so much to be proud of, so much to be optimistic about. I will shortly leave the job and it has been the honor of my life to hold; the second female prime minister - but certainly not the last. I do so with no ill will but with enormous and enduring gratitude to have had the opportunity to serve the country I love.” <em>— Theresa May</em></blockquote><p>(Theresa May ended her speech with tears)</p><p>UK Prime Minister Theresa May announced her resignation as Prime Minister and leader of the country’s Conservative and Unionist Party. Without emotion, I will try to quickly outline my thoughts on this event. (This won’t be a deep article. Please excuse typos too. It came out in a bit of a rush, I must say.):</p><p>Theresa Mary May has served the UK as Prime Minister since 2016. In 2016, a referendum was held to allow the UK leave the EU — aka the Brexit. The results of the referendum and the events that followed form a strong case for current and future political students for generations to come. In summary, the UK has been undergoing negotiations for a deal concerning its relationship with the EU after the Brexit. This has been led by Prime Minister Theresa May and hasn’t gone exactly as planned — leading to her resignation today.</p><p>While this article is about Theresa May’s resignation, let’s take a step back to understand the Brexit. I mean, why does the UK want to leave the EU?!? I feel like giving this to you as an assignment so we keep this story short but I’m also feeling generous today. So I’ll give you a summary:</p><p>In essence, the EU was formed after the World War II to prevent the next war in Europe. (Both world wars began in Europe, come to think of it. ‘Thanks’ to grandmasters Hitler, Mussolini, and Stalin.) The idea was to foster trade between European Countries. So, EU countries would be shooting themselves in the foot if they went to war against each other. That turned out pretty ok. The UK has been a member of the EU since 1973. Being a part of the EU pretty much means acting in the best interest of the union. The only thing is, being in the EU hasn’t exactly favored the Brits. For e.g. the EU’s currency, the Euro, is weaker than the British Pound. And the UK is expected to adopt the Euro as its national currency by 2020 — if it’s still a part of the EU that is. There are also other issues regarding labor, migration and British legislation (which you can research about lol. Let’s keep this article short.)</p><p>Long and short, most Brits haven’t been chill about staying in the EU — or, at least, they weren’t. Despite Former Prime Minister David Cameron’s advocacy, citizens of the UK voted to exit the EU. (David Cameron argued that it was in “the national and economic interest of the UK to stay in a reformed European Union”. When things went south…well he resigned. <a href="https://www.theguardian.com/politics/2016/jun/24/david-cameron-resigns-after-uk-votes-to-leave-european-union">Read more about it here</a>) Cameron wished his successor the very best in negotiating the terms of an exit that benefits the UK. And his successor was Theresa May — who was invited by the Queen to form a new government. Theresa May promised to make Britain a country that works — and when things didn’t work with getting a good Brexit deal, well she announced her resignation in this month of May. In her speech today, Theresa stresses on the importance of compromise in negotiations and indirectly complains about how difficult the parliament house has been in ‘helping’ to negotiate a Brexit deal with the EU.</p><p>The news comes to impact the UK socially, politically and economically. I actually still have to research more into this, so let me not speculate here. But one thing that’s quite obvious is the impact it will have on the British Pound since the news will drive market sentiment. The GPD still remains in the same 3-year freefall in the wake of a likely no-deal Brexit.</p><p>So, what happens next? Well, the Next Prime Minister will continue from where May left off negotiating the terms of the Brexit. It appears that Two-third of British Parliament House already want a no-deal Brexit and since the House will choose the Prime Minister, this could either lead to a no-deal Brexit or a complicated road to organizing another referendum which could eventually result in Brexit.</p><p>If the UK does stay in the EU for some reason by the end of 2020, you might want to find something good to do with your pound notes if you still have them by then. (Archives for your great grandchildren? Good idea! £20,000 of them? Hold that thought?!)</p><p>What a month this has been?! Heightened US-China trade wars and high Brexit emotions. More volatility for the markets I guess.</p><p>Who stands to benefit the most from today’s event? And what are your thoughts on Theresa May’s resignation?</p><p>Watch Theresa May’s resignation here: <a href="https://www.youtube.com/watch?v=_t25xAp270o">https://www.youtube.com/watch?v=_t25xAp270o</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=23d5b1a2c518" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>