<?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 Ruth Fong on Medium]]></title>
        <description><![CDATA[Stories by Ruth Fong on Medium]]></description>
        <link>https://medium.com/@ruthfong?source=rss-2946d3a05431------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*CnE-y12Pt_U3m5Zv70votA.png</url>
            <title>Stories by Ruth Fong on Medium</title>
            <link>https://medium.com/@ruthfong?source=rss-2946d3a05431------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 06:08:07 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ruthfong/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[Applying to the Open Philanthropy AI Fellowship]]></title>
            <link>https://medium.com/@ruthfong/applying-to-the-open-philanthropy-ai-fellowship-48dafe7fe27c?source=rss-2946d3a05431------2</link>
            <guid isPermaLink="false">https://medium.com/p/48dafe7fe27c</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[fellowship]]></category>
            <category><![CDATA[phd]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Ruth Fong]]></dc:creator>
            <pubDate>Fri, 19 Oct 2018 08:58:33 GMT</pubDate>
            <atom:updated>2018-10-20T21:41:49.124Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Update:</strong> I have included links to a few other fellows’ statements.</p><p>A few folks have asked me about applying to the <a href="https://www.openphilanthropy.org/focus/global-catastrophic-risks/potential-risks-advanced-artificial-intelligence/open-philanthropy-project-ai-fellows-program">Open Philanthropy (OP) Project AI Fellows Program</a>, particularly about preparing the personal research statement (quick background: I’m a Ph.D. student at Oxford and was awarded an OP fellowship in 2018).</p><p>Following the example of many researchers I admire who go many extra miles to package and share their <a href="https://github.com/tensorflow/lucid/issues/116">research ideas</a> and <a href="https://www.google.com/search?q=nsf+personal+statement+examples">personal statements</a>, I am writing this short post, sharing <a href="http://ruthcfong.github.io/files/ruth_fong_openphil_ai_fellows_statement_2017.pdf">my personal statement</a>, and linking to a few other fellows’ statements. This is a <strong>hasty</strong> attempt to help anyone preparing for this year’s application cycle (due October 26, sorry!) and democratize the process, particularly for those who have less access to mentors within their networks who have applied to research fellowships.</p><p><strong>This is woefully non-comprehensive. </strong>Unfortunately, I am working towards a research deadline — hence the haste — but I hope to add more to this post after the holidays. If you have further questions, please leave them in the comments. I won’t be able to answer them in time for this year’s deadline but will try to take a stab at them in revision.</p><p><strong>If you’ve successfully applied to research fellowships before,</strong> consider sharing your personal statement and/or writing up your thoughts about the process to share freely.</p><p><strong>If you’re thinking of applying,</strong> please do apply if time permits! Even if you think you’ll be more prepared next cycle and are not a great fit now. I and many others would love to see more diverse and unique fellows.</p><p><strong>For more examples and fellowship advice</strong>, search for “<a href="https://www.google.com/search?q=nsf+personal+statement+examples">NSF personal statement examples</a>” and variants of that query.</p><p><strong>Best of luck!</strong></p><p><strong>DISCLAIMER:</strong> I have no insider knowledge as to what Open Philanthropy is looking for. I am only speaking from my own, limited experience from applying to this fellowship and a few others. These answers are my half-baked intuitive ideas and mine alone.</p><p><strong>Personal Research Statements of a few 2018 OP AI Fellows:</strong></p><ul><li><a href="http://www.stats.ox.ac.uk/~cmaddis/files/maddison_openphil_statement.pdf">Chris Maddison</a></li><li><a href="http://www.foldl.me/uploads/2018/open_phil_research_plan.pdf">Jon Gauthier</a></li><li><a href="http://www.cs.cmu.edu/~noamb/files/open_phil_statement.pdf">Noam Brown</a></li><li><a href="http://ruthcfong.github.io/files/ruth_fong_openphil_ai_fellows_statement_2017.pdf">Ruth Fong</a></li></ul><p><strong>Note: </strong>We are sharing our statements openly so that anyone, not just those in our natural social circles, can have access to them. I want to put an explicit emphasis that a “winning” statement does not have to look at all like one of these.</p><p><strong>What is the Open Philanthropy AI Fellows program?</strong></p><p>“With this program, we seek to fully support a small group of the most promising PhD students in AI and ML who are interested in making the long-term, large-scale impacts of AI a central focus of their research. Fellows are funded through the 5th year of their PhD, and will receive a $40,000 per year stipend, $10,000 per year in research support, and payment of tuition and fees. We encourage applications from 5th-year students, who will be supported on a year-by-year basis; students who will be starting their PhD in Fall 2019; and students with pre-existing funding sources who find the mission and community of the Fellows Program appealing. We are committed to fostering a culture of inclusion, and encourage individuals with diverse backgrounds and experiences to apply; we especially encourage applications from qualified women and minorities.”</p><p>See the below link for more details (all quotes taken from here).</p><p><a href="https://www.openphilanthropy.org/focus/global-catastrophic-risks/potential-risks-advanced-artificial-intelligence/open-philanthropy-project-ai-fellows-program">The Open Philanthropy Project AI Fellows Program</a></p><p><strong>What is a “personal research statement”? (a.k.a. is this a personal statement about my life story or a research proposal?)</strong></p><p>Here’s the official description (emphasis mine):</p><p>“A personal research statement, no more than two pages (excluding references); this document should <strong><em>outline some ideas for research you might conduct</em></strong>. Your research statement is <strong><em>not intended to be a binding description</em></strong>, and we expect that many fellows will end up working on different projects than those they described in their personal research statement.”</p><p>Before proceeding further, let me emphasize that there are many, diverse ways to write a strong statement (there is no singular “winning” formula), and I suspect that that the folks at OP will appreciate many different kinds of statements.</p><p>Personally, on the spectrum of personal statement to research agenda, I would treat this like a “2-page research statement,” based on OP’s description of the personal research statement as non-binding research ideas. This is similar to those for other graduate funding fellowships (i.e., NSF/NDSEG/Hertz or CS-specific ones like NVIDIA/Google/Microsoft; see last question).</p><p><strong>What is a “2-page research statement”?</strong></p><p>2-page research statements typically provide a high-level, clear, and well-motivated description of the research you’d like to conduct in the next <em>N</em> years. This differs from a university application essay and some Ph.D. application essays, which are often more personal and touch on specific research ideas less. This also differs from a more detailed research proposal for a specific project (i.e., grant proposals professors write), which are often longer and more in-depth in terms of proposed methodology, experiments, and background literature.</p><p>It might be daunting to try to describe your research for <em>N</em>=5 years if you are just about to start a Ph.D. program. If that’s the case, I’d recommend considering a more manageable <em>N</em>=2 year time horizon (that’s what I did, but that’s because I needed funding for that duration).</p><p>There are multiple ways to describe research in the 2-pager and, again, there is no one “winning” formula. You could describe a. a few disparate projects you are interested in pursuing, b. one large project, or c. several projects that are loosely related, as just a few options. I personally prefer the last option (c) and aim to have a single, unifying theme that connects and motivates the different ideas in an application essay.</p><p><strong>Who should be my target audience?</strong></p><p>I expect that your application will be read in a similar fashion to how Ph.D. applications are read. In this particular case, because of OP’s focus on machine learning and their <a href="https://www.openphilanthropy.org/focus/global-catastrophic-risks/potential-risks-advanced-artificial-intelligence">grant-making relationships</a> with a few ML groups , I expect the reader to be familiar with machine learning research in some capacity — they could be an expert researcher in a particular area, a well-informed lay person whose follows machine learning development fairly well, or someone working at the intersection of ML + Y (where Y = {government policy, healthcare, public relations/communications, etc.}). I would not expect the reader to be familiar with my particular research area. In light of this, you may want to consider the following: avoid or define highly specific terminology, give a basic overview or providing concise context before diving into details, etc.</p><p><strong>What is the best way to improve my statement/application?</strong></p><p>With days to the deadline, I would prioritize the following:</p><ul><li>Proofreading and revising your statement (ask friends/mentors/letter writers to proofread for you — typos and grammatical errors will distract from even the greatest research ideas).</li><li>Submitting your application on time (including recommendation letters — follow up with your letter writers to make sure they submit on time)</li></ul><p><strong>What is Open Philanthropy looking for?</strong></p><p>I genuinely have no idea. But, if I had to guess, these would be my guesses:</p><ul><li>individuals who demonstrate potential to do interesting, diverse machine learning research</li><li>aspiring researchers who care about the impact of artificial intelligence, particularly its risks, in some capacity</li></ul><p>Usually, fellowships like this one select on <strong>potential</strong>, as they are selecting people at the beginning or early part of the Ph.D. Most of last year’s winners were relatively early (or about to start) their Ph.D. studies and had published a few papers at most (at the time of applying, I was a second year Ph.D. student with one published paper and one paper under review). <strong><em>So apply! Even if you don’t think you’re a great fit now. </em></strong>It’s a helpful exercise to prepare and get feedback on your application (to cast vision for and organize your research dreams). Demonstrating potential can look like a lot of things — for ML research, this may be papers at all levels (i.e., arXiv, workshop, conference, journal), open source code (i.e., GitHub), class projects. My guess is that letters of recommendation are one of the main ways potential is evaluated.</p><p>Given the scope of the OP fellowship (see quote below; emphasis mine), I also estimate that they value <strong>people who, like them, care the future of artificial intelligence, particularly its potential risks, </strong>as it’s applied in various, expected and unexpected ways in the future. Again, “care for AI in the future” can take many forms. For me, this takes a few forms — from my research interest in ML interpretability to my interest in seeing as many diverse people (along numerous axes) as possible enter, succeed in, and influence the field.</p><p>“We expect that AI Fellows will develop many promising research directions beyond those we happen to have listed, and our basic aim is to support <strong><em>excellent researchers who consider potential risks from transformative AI</em></strong> to be a central part of their job, regardless of whether their interests fall into a category we happen to have listed.”</p><p><strong>How does this compare to NSF/NSDEG/Hertz or CS-specific graduate fellowships (i.e., NVIDIA/Google/Microsoft)? How does this affect the statement?</strong></p><p>Quite similarly! The main differences are scale and specificity: NSF awards around <a href="https://www.nsf.gov/news/news_summ.jsp?cntn_id=191361">2000 graduate fellowships</a> each year in a wide, diverse range of scientific fields. Open Philanthropy awarded seven fellowships in its inaugural year (and estimates it will fund 5–10 applicants this year), specific for Ph.D. research in machine learning.</p><p>I usually used the same backbone of a statement and then tailored each statement a smidge. In retrospect, my OP statement was the least, explicitly tailored (most likely because it had the earliest deadline and I subsequently condensed my statement to include more tailored conclusion paragraph). For comparison, here’s <a href="http://ruthcfong.github.io/files/ruth_fong_nvidia_graduate_fellowship_statement_2017.pdf">my statement</a> for the <a href="https://research.nvidia.com/graduate-fellowships">NVIDIA Graduate Fellowship</a> (which I did not receive).</p><p>For OP specifically, if your research interests do not naturally demonstrate “care for the future of AI,” you might want to consider incorporating that somewhere in your application, possibly in your statement. This might look like discussing the motivations and/or impacts of your proposed research.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=48dafe7fe27c" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>