TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

My story as a self-taught AI researcher

364 点作者 emilwallner超过 5 年前

18 条评论

narenst超过 5 年前
This is a really good time to be a Independent Scientist (aka Gentleman scientist) in this field because how nascent deep learning and similar techniques are. It requires a lot of trial and error and time&#x2F;cost investment to bring the AI techniques to the masses.<p>The FAANGs are trying to hire all the top talent (including Emil who wrote the post) but I believe these independent researchers will be the one finding new opportunities to make AI useful in the real world (like colorizing b&amp;w photos, create website code from mockups).<p>The biggest challenge I see for these folks is the access to high quality data. There is a reason Google is releasing so many ML models in production compared to smaller companies. Bridging the data gap requires effort from the community to build high quality open source datasets for common applications.
评论 #22101353 未加载
评论 #22101851 未加载
评论 #22104907 未加载
itsmefaz超过 5 年前
The problem with Emil&#x27;s approach to learning is that it restricts his ability to learn anything that he has no intrinsic goal off. That includes areas like pure mathematics, theoretical computer science, finance, economics, literature, etc. Those subjects require a different sort of motivation than a motivation to just achieve a set goal i.e capitalistic motivation<p>Also, Emil&#x27;s approach to learning will create a flawed sense of expertise. Look at how the article presents him as if he has a <i></i>deep-domain expertise<i></i> which might not be true.<p>One important thing to consider is to look at the article more like content marketing tactic, that FloydHub is using promote its brand which might not serve well for engineers as it lacks some aspect of truth.
评论 #22105593 未加载
评论 #22105467 未加载
K0SM0S超过 5 年前
This was a great read (and great nuggets, like that paper on Intelligence by Chollet).<p>I wonder:<p>— Is math a problem for non-academic researchers?<p>Most papers strike me as requiring a non-trivial knowledge of linear algebra, for instance; and topology sits right behind; the bold seem to take it one up on category theory as we speak, and geometric algebra is quickly gaining traction too. Lots of math, cool math but math nonetheless.<p>Not that you can&#x27;t learn these on your own, but how big is the gap <i>in practice</i>, on the job, compared with actual PhDs in ML&#x2F;math? (how much of a hinderance, a problem it is for the self-taught researcher)<p>— &quot;Contracting&quot; in the field of AI sounds great but, how exactly? Especially solo: what type of clients and how&#x2F;where to find them, what type of &#x27;business proposition&#x27; as a freelancer do you offer, what&#x27;s the pricing structure of such gigs?<p>I mean, I can sell you websites and visuals and stuff, but AI? I know first-hand most SMBs (IME the only real customers for freelancers) are a tough sell: their datasets are tiny and demand scripting skills to sort out (extract business value), not AI, so the value proposition is low for both parties; it&#x27;s still early adoption so 90% don&#x27;t even consider spending 1 cent on &quot;AI&quot; unless as a SaaS (they actually don&#x27;t need to know if it&#x27;s AI or programming).<p>I can imagine tons of fantastic research to do with SMBs, as partners or &#x27;interested sponsors&#x27; (should they reap benefits on a low investment), but really not much yet in the way of &quot;freelancer products&quot; to market and sell for a living. I&#x27;m eagerly anticipating those days, but it&#x27;s more like 2025-2030 as I see it.<p>I would love to hear first hand takes on this.
评论 #22102293 未加载
评论 #22102287 未加载
评论 #22102994 未加载
评论 #22104247 未加载
评论 #22102261 未加载
newswasboring超过 5 年前
I&#x27;m not going to lie, his life story made me jealous. Extremely jealous. He did all the things I wanted to do (speaking in categories, not exact things) and is free to do more. It seems like in some cultures (mine is South Asian) there is a threshold on exploration time. Usually around the age of 28-30 years old (for some even lower than that, I consider myself one of the most fortunate ones). As I approach that number I feel the invisible hand of expectations and responsibilities crushing my spirit. But I must also remember comparisons on life scales don&#x27;t really work and nobody can win neither the happiness Olympics nor the misery Olympics.
评论 #22105698 未加载
评论 #22105419 未加载
评论 #22105570 未加载
评论 #22120933 未加载
octokatt超过 5 年前
Was anyone else really put off by the congratulatory tone of the article, and the #Quirks list on the resume?: <a href="https:&#x2F;&#x2F;github.com&#x2F;emilwallner&#x2F;Emil-Wallner-LinkedIn-Resume#quirks" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;emilwallner&#x2F;Emil-Wallner-LinkedIn-Resume#...</a>
评论 #22104574 未加载
评论 #22104890 未加载
评论 #22103365 未加载
qntty超过 5 年前
&quot;Many are realizing that education is a zero-sum credential game.&quot;<p>Can this silly meme die already? Maybe it&#x27;s understandable coming from an economist who values education for no other reason than it&#x27;s economic effects, but it&#x27;s strange coming from someone who clearly understands the value of personal development.
评论 #22101888 未加载
评论 #22101777 未加载
评论 #22101836 未加载
评论 #22104050 未加载
评论 #22101632 未加载
exdsq超过 5 年前
Survivorship bias or reality:<p>3 months learning FastAI, 3-12 months personal projects and consulting, 2 months flashcards of ~100 papers, 6 months to publish a paper<p>What does he mean by ‘paper’? A Medium post? NeurIPS?
评论 #22104632 未加载
评论 #22105324 未加载
bluetwo超过 5 年前
The thing that disappoints me about the aspirations of being a researcher is that the goal is to get paid to study AI, not solve real-world problems.<p>I would rather build a small company by solving a real problem than work for a big company spinning my wheels.
评论 #22101775 未加载
评论 #22101819 未加载
ineedasername超过 5 年前
I think in these sorts of discussions two concepts with the same name tend to get conflated, so I think it&#x27;s important to make a distinction between:<p>1) <i>AI Research</i> as applying&#x2F;tweaking known ML&#x2F;DL methods to a novel problem. I would term these something like &quot;AI Engineering Research&quot;<p>2) <i>AI Research</i> as examining the theoretical frameworks &amp; approaches to ML&#x2F;DL in a way that may itself lead to shifts in the understanding of ML&#x2F;DL as a whole and&#x2F;or develop fundamentally new tools for the purpose of #1. What might be termed &quot;basic&quot; or &quot;pure&quot; research.<p>I&#x27;m not placing one of these above the other in terms of importance. They are both necessary, and they form a virtuous feedback loop between the two that, one without the other, would see the other wither on the vine.<p>In the example of this particular person, Emil Wallner, he appears to be doing #1, and perhaps doing so in a way that might help inform more of #2.
评论 #22102844 未加载
评论 #22102985 未加载
rmah超过 5 年前
Is this guy actually a <i>researcher</i> in the way most people would think of it? That is, someone who pushes the boundaries of science; who develops new AI techniques or finds the hard boundaries of existing AI techniques; who finds new ways compose multiple AI techniques cohesively; who explores the theoretical foundations of AI.<p>Or is he someone who uses AI techniques to solve problems (and then wrote a paper about it)? I can&#x27;t help but wonder a bit.
评论 #22102645 未加载
评论 #22103614 未加载
评论 #22102717 未加载
评论 #22102935 未加载
wigl超过 5 年前
This reeks of survivorship bias to me. I much prefer Andreas Madsen&#x27;s more sober and self-conscious take on independent research [0].<p>&gt; I’d spend 1-2 months completing Fast.ai course V3, and spend another 4-5 months completing personal projects or participating in machine learning competitions... After six months, I’d recommend doing an internship. Then you’ll be ready to take a job in industry or do consulting to self-fund your research.<p>Where are these internships that will hire you based on your completion of Fast.ai (if done in 1-2 months by a beginner I assume it&#x27;s only part 1) alone, especially in 2020? How many are going to place in a Kaggle competition with just half a year of experience? More importantly, just how many people are privileged&#x2F;secure enough to put their all into learning, with no sense of security or peer support?<p>&gt; I started working with Google because I reproduced an ML paper, wrote a blog post about it, and promoted it. Google’s brand department was looking for case studies of their products, TensorFlow in this case. They made a video about my project. Someone at Google saw the video, though my skill set could be useful, and pinged me on Twitter.<p>So what really mattered was self-promotion, good timing, and luck.<p>&gt; Tl;dr, I spent a few years planning and embarking on personal development adventures. They were loosely modeled after the Jungian hero’s journey with the influences of Buddhism and Stoicism.<p>Why does the author have to present his life like one would in a fucking college essay?<p>[0] <a href="https:&#x2F;&#x2F;medium.com&#x2F;@andreas_madsen&#x2F;becoming-an-independent-researcher-and-getting-published-in-iclr-with-spotlight-c93ef0b39b8b" rel="nofollow">https:&#x2F;&#x2F;medium.com&#x2F;@andreas_madsen&#x2F;becoming-an-independent-r...</a>
评论 #22102541 未加载
评论 #22102919 未加载
评论 #22105193 未加载
vector_spaces超过 5 年前
&gt; Creating value with your knowledge is evidence of learning. I see learning as a by-product of trying to achieve an intrinsic goal, rather than an isolated activity to become educated.<p>&gt; Early evidence of practical knowledge often comes from usage metrics on GitHub, or reader metrics from your work blog. Progress in theoretical work starts by having researchers you consider interesting engage with your work.<p>&gt; Taste has more to do about character development than knowledge. You need taste to form an independent opinion of a field, having the courage to pursue unconventional areas and to not get caught up in self-admiration.<p>When I study abstract interpretation or lattices, I&#x27;m doing so because I find those subjects interesting and beautiful, and studying math relaxes me. I can lie to myself and say that it&#x27;s improving my problem solving ability and that it&#x27;s like doing mental yoga and will make me better at my job or some baloney, but that&#x27;s not why I do it.<p>I can spend time with a plant in my garden, take a cutting, root it and replant it, and watch it grow, learn the ebbs and flows of its watering needs through the seasons, learn what its seed pods look like, and eventually watch it die through some misstep of my own or otherwise.<p>And in doing so, I am learning, and building a mental model for this plant and an intuition for it, but I&#x27;m not &quot;creating value&quot; in some weird capitalist sense, which I feel always underlies these sorts of opinions about learning and education, and people who self-identify as &quot;makers&quot; in general. It rubs me the wrong way because it encourages a very narrow view of the human experience and what it means to learn and why we should learn.
评论 #22105294 未加载
anentropic超过 5 年前
I think most of us here on this site could do the same in terms of learning and research<p>Seems to me the difficult part is how to support yourself financially while spending your time doing interesting learning and research, or how to get paid to do it<p>Maybe the most important detail in the story is &quot;He co-founded a seed investment firm that focuses on education technology&quot; but it is not discussed further
LemonAndroid超过 5 年前
I don&#x27;t see how this is self-taught, as the person got picked up for an internship and could learn from experts first-handly.<p>FAKE.
评论 #22104884 未加载
NWM123超过 5 年前
I personally found this article to be very interesting. I don&#x27;t know much about AI, but I was fascinated by the discussion of peer to peer educational system. I believe that they will become more prevalent as student loan payments cause debt to so much of our population in order to get an education .
jshowa3超过 5 年前
I don&#x27;t know why people think getting a credential does nothing or that people &quot;copy and paste&quot; the assignments. Sure it may be possible, but what prevents people from copying and pasting public git repos?<p>Either way, this whole focus on &quot;portfolios are everything and credentials are meaningless&quot; spits in the face of all the work I did to get my university education. And it didn&#x27;t involve &quot;copying assignments&quot;. And you come out with one hell of a portfolio if you take your education seriously.<p>I mean I don&#x27;t think self-educated people are without merit. I happen to think they&#x27;re really important. But I only ever see them rag on higher education, despite them having &quot;never been there&quot;.<p>Just another example of wunderkin super genius knows all because he was able to follow a non-standard path and make it. Glad he was smart enough to become a Google employee. But I question whether he should be giving advice on paths to get there when there&#x27;s always many paths to a position. And especially after reading his brief comments on how credentials imply you&#x27;re a liar.
评论 #22103799 未加载
评论 #22103874 未加载
评论 #22103762 未加载
ptah超过 5 年前
&gt; deep learning internship at FloydHub.<p>nice to be able to work for free and not starve
DoctorOetker超过 5 年前
This is a great example of how we &quot;collectively&quot; [1] conflate phenomena, skillsets, ... into one topic: machine learning, AI.<p>1) There is the general phenomena or collective project, where hardware, algorithms and human insights are improved to approach the situation of man-made intelligent machines.<p>2) There are the people who are designing algorithms, using mathematical intuition and knowledge, analogies with physics, etc... Most people would agree these people are doing optimization &#x2F; machine learning &quot;proper&quot;.<p>3) There are the people working on improving hardware for machine learning &#x2F; optimization purpouses, by looking at the most performant algorithms, breaking them down into primitive operations and requirements for hardware, there are also people working on the algorithms themselves and finding computational shortcuts (which can end up in software or hardware, can end up as proprietary knowledge or common knowledge, ...). The distinction between hard and software is somewhat blurry, since hardware designers can optimize or implement a section of software into hardware. A lot of this can still be considered ML &quot;proper&quot;.<p>4) Then there are the people who apply the ML frameworks and their exposed choices and settings to a specific problem domain. Many of them don&#x27;t need to understand the internals if they don&#x27;t need state of the art results. Many would nevertheless benefit from understanding the internals, and the requisite math. What I propose is to stop calling their activity as Machine Learning, and instead call it Machine Teaching. They are teachers, and just like elite schools they can choose which specific type of available student they will teach, and they can tweak (or filter from a large family of students) which student they select to teach the task at hand. There are bound to be many advantages of having actual human teachers get involved in machine teaching. These people will not be proficient in designing novel families of students unless they also know the requisite math, and identify those ML papers that are ML &quot;proper&quot; instead of ML &quot;teacher&quot;. When trying to find important foundational insights in ML &quot;proper&quot; one is typically overwhelmed by a large surplus of ML &quot;teacher&quot; type papers. These are important datapoints, and necessary to advance human insight into ML &quot;proper&quot;, but they are data, not knowledge. There are actual ML &quot;proper&quot; knowledge papers out there that explain why a certain phenomena is such and so, and they get very little attention because they necessarily lag the breakthrough ML datapoint paper, and most ML &quot;teachers&quot; don&#x27;t have the math background to understand them. So the probability that a given ML &quot;proper&quot; researcher <i>fundamentally</i> improves the state of the art is much higher than the probability that a given ML &quot;teacher&quot; will <i>fundamentally</i> improve the state of the art. At the same time the probability that a given <i>fundamental</i> breakthrough was achieved by an ML &quot;teacher&quot; is higher than the probability that a given <i>fundamental</i> breakthrough was achieved by an ML &quot;proper&quot; researcher:<p>P( Breakthrough | Proper ) &gt; P ( Breakthrough | teacher)<p>while<p>P ( Teacher | Breakthrough ) &gt; P ( Proper | Breakthrough )<p>Since most people don&#x27;t have the broad math &#x2F; physics &#x2F; ... knowledge to draw on, the number of ML &quot;teachers&quot; is much higher than ML &quot;proper&quot; researchers.<p>[1] well, really, some actors have vested interests in conflating those together...<p>EDIT: just to be clear, I am not complaining about ML Teachers, we need the ML Teachers, and their breakthrough datapoints. What I am complaining about, is conflating both activities of ML Proper and ML Teaching. This makes it harder for the few ML Proper researchers to find each other&#x27;s insights.