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。

Adventures in Improving AI Economics

249 点作者 oliver101将近 5 年前

12 条评论

eanzenberg将近 5 年前
In my experience, there are just a lot of &quot;bad&quot; AI&#x2F;ML engineers who don&#x27;t fundamentally understand what data can do, what ML algorithms can handle, and how to piece it together to produce something of value to the end user. A couple of these people on a team can torpedo a project. Worse are those who sabotage projects or are general pain points of hindering progress. These may be jaded people who don&#x27;t believe that ML has any value yet have titles like Data Scientist or ML engineer, and can bring team morale down. The economics are similar to a grad-school research project, yet is infiltrated by all sorts of people with 3 month certificates believing they are the star of the show.<p>The most important element of AI project success is the right people and the right team. Projects are long-term and failure can be often. It&#x27;s not easy to succeed but cultivating the right people and their mindset is in my opinion a needle mover for AI projects, more-so than what data is available, what algorithms are tried, and what shiny framework people want to use.
评论 #24155159 未加载
评论 #24152295 未加载
评论 #24151332 未加载
评论 #24152373 未加载
alextheparrot将近 5 年前
a16z has a podcast where they explored gross margins a month back. The panel called out AI as an example of a software business that has a high likelihood of not having standard SaaS margins (Most of the panel thought this could be a limitation).<p>The podcast is nice because I think it holistically explores gross margins in a way that you start to understand how it might impact AI as a viable primary business model and valuations related to companies who that is the case for. Quite complementary to the article.<p>Might be interesting to people who are interested in this article: <a href="https:&#x2F;&#x2F;open.spotify.com&#x2F;episode&#x2F;79lJCrHB3nBn1qXCxKA5s7?si=RFlHPM0_QwWHZsqwf3nbxA" rel="nofollow">https:&#x2F;&#x2F;open.spotify.com&#x2F;episode&#x2F;79lJCrHB3nBn1qXCxKA5s7?si=R...</a>
评论 #24152300 未加载
评论 #24150478 未加载
motohagiography将近 5 年前
Good analysis and great of them to share their thinking. Does feel like this could have been a tweet that said the necessary condition for successful ML solution is applying it to a problem that has asymmetric upside.<p>Great for telling people they should get tested for diseases, terrible for diagnosis. In the alerting first case, consequences of being wrong are no better than base rate as they wouldn&#x27;t have been tested otherwise, and the upside saves a life. In the latter diagnosis case, the consequences of being wrong are catastrophic, and it is substituting for the best available judgment. Similarly, it&#x27;s great for fraud detection, terrible for making credit decisions, because the false negative rate is essentially externalized. It&#x27;s good for finding opportunities, bad for providing services. So funnels and conversion pipelines it&#x27;s great for.<p>So perhaps there&#x27;s an ironic Turing test for ML solutions that is related to the relationship between the size of a group of people and the effect of mean reversion of their collective intelligence on their behaviour makes them indifferent to the perceived intelligence of the model, whereas a given individual will find the results of the model unsatisfying. From an indifference perspective, AI can fool some of the people all the time, and all the people some of the time, but no confusion matrix satisfies all the people all the time. Economically, ML will be useful for creating simple and cheap services that people who can&#x27;t afford better will use, and substitute up from them when they can afford better, known as &quot;inferior goods.&quot; There may be a hard limit on ML providing &quot;normal goods,&quot; to individuals at scale for this reason. Lots of money to be made, but lots to be wasted tweaking your ROC curve to in the hope of creating a normal good.<p>I yell from the rooftops every chance I get that &quot;the confusion matrix is the product.&quot; That is, your FP&#x2F;FN&#x2F;TP&#x2F;TN rate is your product, and you are optimizing your system for the weights your customer assigns to those variables.<p>There is another ML&#x2F;DL use case I&#x27;m hacking on that is about enabling privacy, but even this reduces to the asymmetry of the upside&#x2F;downside of the confusion matrix. Obviously the article is more nuanced than this, but I think this heuristic is a key tool for reading articles like it.
评论 #24151759 未加载
评论 #24152345 未加载
oliver101将近 5 年前
&gt; This is the crux of the AI business dilemma. If the economics are a function of the problem – not the technology per se – how can we improve them?<p>The article focusses on the costs of resources to build a model (annotated data + compute) but the economics are also affected by the ongoing cost of making a prediction error. False positives and false negatives usually have a different cost and each user might have their own preferences:<p>e.g. &quot;show me all the content that&#x27;s a bit relevant&quot; vs &quot;show me just the content that&#x27;s really relevant&quot;.<p>If you can write out the loss function in $$$ terms not just accuracy, then you&#x27;re closer to either abandoning the problem or finding a profitable AI model.
评论 #24157090 未加载
tosh将近 5 年前
this is more related to the previous a16z article on the topic but I found &quot;Data as a Service&quot; by Auren Hoffman a great read for thinking about businesses that sell access to machine learning models<p><a href="https:&#x2F;&#x2F;www.safegraph.com&#x2F;blog&#x2F;data-as-a-service-bible-everything-you-wanted-to-know-about-running-daas-companies" rel="nofollow">https:&#x2F;&#x2F;www.safegraph.com&#x2F;blog&#x2F;data-as-a-service-bible-every...</a>
zamadatix将近 5 年前
&quot;Andreessen Horowitz (known as &quot;a16z&quot;) is a venture capital firm in Silicon Valley, California&quot;<p>In case anyone was as confused as I was about what a16z means - it&#x27;s just the company not a new abbreviated term related to AI.
评论 #24150141 未加载
评论 #24150590 未加载
评论 #24150617 未加载
评论 #24150454 未加载
PaulHoule将近 5 年前
Gr8 article.<p>I&#x27;d add that caveat that software dev processes can be well controlled or not well controlled. AIML is not so much a new kind of project but it is a project likely to be poorly controlled.<p>Another thing they don&#x27;t mention is that AIML projects break the agile assumption that you can manage with only punchclock, not calendar time.<p>Imagine you have a 2 week sprint and it takes 1 week to train a model. You have to get the training started in the first week, and any tasks that need to be done to start training have to start before that.<p>This of course means applying PERT chart thinking even if you don&#x27;t make PERT charts. It often isn&#x27;t that hard but in an agile shop that mistakes the map for the territory they will start the 1 week job consistently on the last day of the sprint.<p>The &#x27;containerization&#x27; process they describe is close to the methods used by East coast defense contractors (in a band between research triangle park and the applied physics dept at John Hopkins in baltimore) to get high accuracy. Also they were what IBM Watson did as opposed to what people thought they did.<p>It&#x27;s amazing those methods have remained so obscure, but the mind that is impressed with BERT is going to be impervious to asymtopes. That article should be telling people to run not walk away from those kind of models -- it is how you always be a bridesmaid but never a bride.
mensetmanusman将近 5 年前
Good analogy about discovery of Pharma molecules.<p>It’s really fun to think about the fact that Tesla has more than enough data to unlock autonomous vehicles, but all that is missing is the correct AI architecture to get it working...<p>Who will figure out how to code that? Will it be a breakthrough, or can sub-optimal architectures eventually reach equilibrium with 10x or 100x the amount of time&#x2F;data processing.
评论 #24150400 未加载
评论 #24150107 未加载
评论 #24149919 未加载
mijail将近 5 年前
I&#x27;m happy a VC is providing these insights. If the economics of AI don&#x27;t make sense by helping increase profits or cutting costs... it&#x27;s going to be a long road to reaching the &quot;promised land.&quot;<p>I&#x27;m biased but in a lot of industries synthetic data has the potential to balance the costs from the perspective of data acquisition and preparation as well as model testing.<p>This article doesn&#x27;t focus too much on the edge side of things but one pattern I&#x27;m seeing is that edge deployment can be notoriously resource intensive and time consuming.
gk1将近 5 年前
One way to deal with AI&#x2F;ML shortcomings I&#x27;ve seen is to require end-user intervention for edge cases; such as a support chatbot that transfers the customer to a human rep if it can&#x27;t understand the issue. Human intervention isn&#x27;t mentioned in the article but maybe they&#x27;d put that under &quot;narrow the problem,&quot; or they may not consider that a solution since human involvement eats into margins.<p>I believe all software companies will be AI companies in &lt;5 years. By then, not having AI&#x2F;ML would be like not having a database today. There will be no choice but to deal with the long tail, and the competitive advantage will go to the company that does it better. That makes this advice all the more timely and important, and it also means opportunities for startups to innovate in this space. Eg, better model optimization, low-cost operations without regressing to colocated GPUs, etc.<p>&quot;The critical design element is that each model addresses a global slice of data... There is no substitute, it turns out, for deep domain expertise.&quot; Totally true for marketing as well. Much more effective to define audience segments and tailor the messaging and marketing for each.
known将近 5 年前
<a href="https:&#x2F;&#x2F;yts.mx&#x2F;movies&#x2F;robot-frank-2012" rel="nofollow">https:&#x2F;&#x2F;yts.mx&#x2F;movies&#x2F;robot-frank-2012</a> show subtle issues related to AI in real life;
tigerbelt将近 5 年前
Indubitably