A very fun jaunt through the early days of front-end web development. They are open to pull requests, so submit one if you’re into this kind of fun chronicling of our weird history!
That CSS3 Button generator really hits home. 😬
Direct Link to Article — Permalink
The post Dark Ages of the Web appeared first on CSS-Tricks.
You can support CSS-Tricks by being an MVP Supporter.
from CSS-Tricks https://pavellaptev.github.io/web-dark-ages/
Never miss an insight. We’ll email you when new articles are published on this topic.
from McKinsey https://www.mckinsey.com/business-functions/strategy-and-corporate-finance/our-insights/five-fifty-the-quickening
Here lies one from a distant star, but the soil is not alien to him, for in death he belongs to the universe. ― Clifford D. Simak, Way Station
Just a few weeks ago, the European Southern Observatory’s Very Large Telescope (ESO’s VLT) released an image of a nascent solar system being born. This observatory was the first to ever image an exoplanet, released in 2004.
The SPHERE instrument on the VLT blocks light from distant stars using a device called a coronagraph, allowing faint planets to be seen in images recorded by the telescope. The video below shows a look at this new discovery from the European Southern Observatory’s Very Large Telescope.
However, this technique produces an aura of light, like that caused by blocking the Sun with a finger. Fortunately, light emitted from stars is unpolarized — electromagnetic waves the oscillate randomly in different directions. Once it strikes a surface and reflects, light becomes partial polarized. This allows researchers to use a technique similar to polarized sunglasses to see detail in the images.
Older planets, like those found in our own solar system, have cooled down too far to be found by alien astronomers using this technique. However, younger planets are warm, allowing the planets to show via infrared radiation.
Adaptive optics eliminate the effects of the atmosphere on light from space, providing astronomers on the ground with images having clarity rivaling those taken by space telescopes.
“SPHERE is trying to capture images of the exoplanets directly, as though it were taking their photograph. SPHERE can also obtain images of discs of dust and debris around other stars, where planets may be forming. In either case, direct imaging is extremely hard to do,” researchers wrote on the ESO website.
Analysis of the study was published in the Astrophysical Journal Letters.
Astronomers believe this new study will assist astronomers in learning more about our own solar system and other families of planets with stars like our own.
This article was originally published on The Cosmic Companion by James Maynard, founder and publisher of The Cosmic Companion. He is a New England native turned desert rat in Tucson, where he lives with his lovely wife, Nicole, and Max the Cat. You can read this original piece here.
Astronomy News with The Cosmic Companion is also available as a weekly podcast, carried on all major podcast providers. Tune in every Tuesday for updates on the latest astronomy news, and interviews with astronomers and other researchers working to uncover the nature of the Universe.
from The Next Web https://thenextweb.com/syndication/2020/07/27/look-human-its-the-first-photo-of-planets-orbiting-a-sun-300-light-years-away/
Posted by Cyrus-Shepard
There’s a new ranking factor in town: Core Web Vitals. Expected in 2021, this Google-announced algorithm change has a few details you should be aware of. Cyrus Shepard dives in this week on Whiteboard Friday.
Click on the whiteboard image above to open a high-resolution version in a new tab!
Howdy, Moz fans. Welcome to another edition of Whiteboard Friday. I’m Cyrus Shepard here at Moz. Today we’re talking about the next official Google ranking factor — Core Web Vitals. Now what do I mean by official ranking factor?
Google makes hundreds of changes a year. Every week they introduce new changes to their algorithm. Occasionally they announce ranking factor changes. They do this in particular when something is important or they want to encourage people, webmasters to make changes to their site beforehand. They do this for important things like HTTPS and other signals.
So this is one they actually announced. It’s confusing to a lot of people, so I wanted to try to demystify what this ranking signal means, what we can do to diagnose and prepare for it, and basically get in a place where we’re ready for things to happen. So what is it? Big first question.
So these are real-world experience metrics that Google is looking at, that answer things like: How fast does the page load? How fast is it interactive? How fast is it stable? So basically, when visitors are using your web page on a mobile or a desktop device, what’s that experience like in terms of speed, how fast can they interact with it, things like that.
Now it’s joining a group of metrics that Google calls Page Experience signals. It’s not really a standalone. It’s grouped in with these Page Experience metrics that are separate from the text on the page. So these are signals like mobile friendliness, HTTPS, intrusive interstitials, which are those pop-ups that come on and appear.
It’s not so much about the text of the page, which are traditional ranking signals, but more about the user experience and what it’s like, how pleasant it is to use the page, how useful it is. These are especially important on mobile when sometimes the speed isn’t as high. So that’s what Google is measuring here. So that’s what it is.
Well, it’s going to affect all regular search results, mobile and desktop, based on certain criteria. But also, and this is an important point, Core Web Vitals are going to become a criteria to appear in Google Top Stories. These are the news results that usually appear at the top of search results.
Previously, AMP was a requirement to appear in those Top Stories. AMP is going away. So you still have to meet the requirements for regular Google News inclusion, but AMP is not going to be a requirement anymore to appear in Top Stories. But you are going to have to meet a minimum threshold of Core Web Vitals.
So that’s an important point. So this could potentially affect a lot of ranking results.
Well, Google has told us that it’s going to happen sometime in 2021. Because of COVID-19, they have pushed back the release of this within the algorithm, and they want to give webmasters extra time to prepare.
They have promised us at least six months’ notice to get ready. As of this recording, today we have not received that six-month notice. When that updates, we will update this post to let you know when that’s going to be. So anytime Google announces a ranking factor change, the big question is:
How much do I have to worry about these metrics, and how big of results are we going to see shift in Google SERPs? Well, it’s important to keep in mind that Google has hundreds of ranking signals. So the impact of any one signal is usually not that great. That said, if your site is particularly poor at some of these metrics, it could make a difference.
If you’re in a highly competitive environment, competing against people for highly competitive terms, these can make a difference. So it probably is not going to be huge based on past experience with other ranking signals, but it is still something that we might want to address especially if you’re doing pretty poorly.
The other thing to consider, some Google signals have outsized impact beyond their actual ranking factors. Things like page speed, it’s probably a pretty small signal, but as users experience it, it can have outsized influence. Google’s own studies show that for pages that meet these thresholds of Core Web Vitals, visitors are 24% less likely to abandon the site.
So even without Core Web Vitals being an official Google ranking factor, it can still be important because it provides a better user experience. Twenty-four percent is like gaining 24% more traffic without doing anything, simply by making your site a little more usable. So even without that, it’s probably still something we want to consider.
So I want to jump briefly into the specifics of Core Web Vitals, what they’re measuring. I think people get a little hung up on these because they’re very technical. Their eyes kind of glaze over when you talk about them. So my advice would be let’s not get hung up on the actual specifics. But I think it is important to understand, in layman’s terms, exactly what’s being measured.
More importantly, we want to talk about how to measure, identify problems, and fix these things if they happen to be wrong. So very briefly, there are three signals that go into Core Web Vitals.
The first being largest contentful paint (LCP). This basically asks, in layman’s terms, how fast does the page load? Very easy concept. So this is hugely influenced by the render time, the largest image, video, text in the viewport.
That’s what Google is looking at. The largest thing in the viewport, whether it be a desktop page or a mobile page, the largest piece of content, whether it be an image, video or text, how fast does that take to load? Very simple. That can be influenced by your server time, your CSS, JavaScript, client side rendering.
All of these can play a part. So how fast does it load?
The second thing, cumulative shift layout (CSL). Google is asking with this question, how fast is the page stable? Now I’m sure we’ve all had an experience where we’ve loaded a page on our mobile phone, we go to click a button, and at the last second it shifts and we hit something else or something in the page layout has an unexpected layout shift.
That’s poor user experience. So that’s what Google is measuring with cumulative shift layout. How fast is everything stable? The number one reason that things aren’t stable is that image sizes often aren’t defined. So if you have an image and it’s 400 pixels wide and tall, those need to be defined in the HTML. There are other reasons as well, such as animations and things like that.
But that’s what they’re measuring, cumulative shift layout.
Third thing within these Core Web Vitals metrics is first input delay (FID). So this question is basically asking, how fast is the page interactive? To put it another way, when a user clicks on something, a button or a JavaScript event, how fast can the browser start to process that and produce a result?
It’s not a good experience when you click on something and nothing happens or it’s very slow. So that’s what that’s measuring. That can depend on your JavaScript, third-party code, and there are different ways to dig in and fix those. So these three all together are Core Web Vitals and play into the page experience signals. So like I said, let’s not get hung up on these.
Let’s focus on what’s really important. If you have a problem, how do you measure how you’re doing with Core Web Vitals, and how do you fix those issues? Google has made it very, very simple to discover. The first thing you want to do is look in Search Console. They have a new report there — Core Web Vitals. They will tell you all your URLs that they have in their index, whether they’re poor, needs improvement, or good.
If you have URLs that are poor or needs improvement, that’s when you want to investigate and find out what’s wrong and how you can improve those pages. Every report in Search Console links to a report in Page Speed Insights. This is probably the number-one tool you want to use to diagnose your problems with Core Web Vitals.
It’s powered by Lighthouse, a suite of performance metric tools. You want to focus on the opportunities and diagnostics. Now I’m going to be honest with you. Some of these can get pretty technical. You may need a web developer who is an expert in page speed or someone else who can comfortably address these problems if you’re not very technical.
We have a number of resources here on the Moz Blog dealing with page speed. We’ll link to those in the comments below. But generally, you want to go through and you want to address each of these opportunities and diagnostics to improve your Core Web Vitals score and get these out of poor and needs improvement into good. Now if you don’t have access to Search Console, Google has put these reports in many, many tools across the web.
Lighthouse, of course, you can run for any page. Chrome Dev Tools, the Crux API. All of these are available and resources for you to find out exactly how your site is performing with Core Web Vitals and go in and we have until sometime in 2021 to address these things. All right, that’s it.
That’s Core Web Vitals in a nutshell. We’ve got more than six months to go. Get ready. At least at a very minimum dive in and see how your site is performing and see if we can find some easy wins to get our sites up to speed. All right. Thanks, everybody.
Video transcription by Speechpad.com
Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don’t have time to hunt down but want to read!
from Moz https://moz.com/blog/core-web-vitals
We’re living in a very interesting time. New inventions are being created. New jobs and lifestyles become possible. The very definition of the word “work” changes.
A month ago, I argued that the job of the future is content creation.
It’s true.
Something similar happened when we switched from physical labor employment to intellectual labor.
About one hundred years ago, we stopped going to factories and started working in offices. We didn’t need that much physical labor anymore because most of it was automated. Intellectual work and the post-industrial economy soared.
Now, we’re in a new transition. There are almost eight billion people on this planet — and while there is still a large percentage of those who don’t have access to the internet (about half, to be exact) — the western, developed world just doesn’t need that many employees.
This is what the 2008 real estate crisis showed us. There were many layoffs, but most of them didn’t recover. Why? The CEOs were secretly happy that they had to fire all those people. The brutal truth is that they just didn’t need them.
This COVID crisis also showed us just that.
Most employees were either forced to work from home or go on a three-month-long vacation. The world economy tanked for a bit (then recovered to all-time-highs), but not because all those people couldn’t work — but because most offline businesses had to shut down.
There was less demand for products and services.
That last point is crucial.
What many people still don’t realize is that we live in a “demand-driven” economy. In this economy, bitcoin grows not because it’s backed by some incredible asset (it’s not backed by anything), but simply because it grows. It grows because people buy it. And when people see that it grows, they buy some more of it. The rest of the economy is built in a similar way.
The whole world economy revolves around the demand of 1% of the global population.
And because the population grows (especially fast in the developed countries), you have demand that’s accelerating.
More demand = more opportunities for business. And less need for work. In the conventional sense, at least.
This is where global concepts clash with personal demands.
“But how do I feed myself then, if nobody needs my work?” you might ask.
I am glad you did.
You have three choices.
Choices 1–2: you can enter one of the two economies:
Choice 3: become an entrepreneur and distribute value in some other way — mainly by organizing people from the ‘gig economy’ to create something that has demand from consumers.
And while yes, most people are still working as full-time employees, the old concept of “career” is becoming quickly obsolete — much like the estates and ranks of the medieval ages.
In the next decade or two, we’ll see that these three choices are all you’ve got.
You’re probably not surprised to see the words “gig economy” — as it existed for quite some time. After the COVID pandemic ceases, we’ll see more of it booming, as ex-employees go freelance. You’re also probably not surprised to see entrepreneurship in this, as the media is full of stories of new-born small business owners. Both of these choices are the by-product of our demand-driven economy.
But the words ‘passion economy’ are probably new to you.
That’s because this economy is new, the youngest of the three. It’s nevertheless quite real.
There is a guy who lives on a farm and feeds his animals. Just twenty years ago, he’d be considered as a slacker and hermit. Today he’s a YouTuber, making more than $100K/yr from Patreon subscriptions.
There are writers on this platform without college degrees who work two hours per day and make the same amount of money as someone who went to an Ivy League school and busted their asses off climbing the career ladder.
This economy — whether you call it ‘creative economy’ or ‘passion economy’ or plain-old ‘influencers’ — existed for a decade. It started off with pioneers and legends like Casey Neistat, Joe Rogan, Tim Ferriss, the Kardashians, and others — but now it’s becoming mainstream.
Today we see more and more people making not millions — but solid full-time incomes by just being themselves.
In a way, we’re going back to the roots: some 80,000 years ago, our ancestors spent 99.9% of the time scratching their dicks, and at night they told stories by the campfire.
Now we’re circling back. As the world becomes safer, the population grows, demand grows, more people become redundant and conventional work becomes obsolete, we’ll see a whole new class of professions created, which will allow humans to tell stories for a living.
What else should they do?
Hans Moravec wrote, “In time, almost all humans may work to amuse other humans, while robots run competitive primary industries, like food production and manufacturing.”
We’re accustomed to thinking of such prophecies as sci-fi or things of the distant future, which we might not live to see.
But it’s already here.
Take me, for example. I am 22. If I was born just ten years earlier, I would have been miserable and without a job. Or I would have worked as a journalist for a shitty Russian newspaper, and that’s all. My skills, desires, ambitions would have been unnecessary.
Today though, because we have platforms like Medium, YouTube, Spotify — I can go straight to creating art for a living. I don’t have to take on a “day job” (a term coined in the twentieth century by struggling musicians), I can get paid the instant I create something valuable, interesting, new.
Not only careers become obsolete, but degrees themselves too. It used to be that a degree was a platform. It was something that gave people opportunities: to find like-minded people, to have credibility, to be seen, and hired. Today it’s nothing. When people ask me where I went to college, I usually reply, “Babson”, although I never finished school, and nobody cares.
But this doesn’t mean you don’t need a platform. It just changed form.
The platform of this day and age is the audience. Your people. Network. Its trust.
And just like a good college degree, it can give you a brand, credibility, and — what’s more important — community. You can have $0 in the bank, but if you have an engaged audience that trusts you — you’ll survive.
They say in Medium’s curation guidelines that every article should have a strong CTA (call-to-action) at the end. So here’s mine.
If there’s one thing you take away from this thought piece, let it be this: the world has changed.
You don’t have to live by the old rules. Don’t listen to your parents and don’t trust people over 35 — most of them don’t know what they’re talking about — they were born and raised in another world.
You have three economies to choose from. Working in an office full-time is not one of them — and is a recipe for disaster over the long-term (although now it might be a lucrative option).
But most importantly, what this new world has brought us is power — the ability to choose.
So choose yourself.
The Job of the Future Is Content Creation
Soon We’ll Be Creating Content For a Living was originally published in The Startup on Medium, where people are continuing the conversation by highlighting and responding to this story.
from The Startup – Medium https://medium.com/swlh/soon-well-be-creating-content-for-a-living-b7afdb942f39?source=rss—-f5af2b715248—4
As the United States experiences a coin shortage due to the pandemic, some brick-and-mortar retailers are forcing consumers to tell cashiers to “keep the change.”
Kroger is the largest chain so far to do so, temporarily refusing to give coins out as change. Customers can either round up their bills to the nearest dollar and give the excess to charity or get their change stored on loyalty cards.
“The Federal Reserve is experiencing a significant coin shortage that is impacting our store operations and ability to provide change,” the company wrote on Twitter. “As a result, the company is implementing a new process for providing change to customers. In all staffed lanes, coin change owed to the customer can be applied to your loyalty card and can be used on your next in-store, Pick-Up or Delivery purchase. Alternately, we can round your transaction up to the nearest dollar and donate it to your local foodbank.”
The Fed had said in a June 11 statement that the pandemic “has significantly disrupted the supply chain and normal circulation patterns for U.S. coin. In the past few months, coin deposits from depository institutions to the Federal Reserve have declined significantly, and the U.S. Mint’s production of coin also decreased due to measures put in place to protect its employees.”
Kroger isn’t alone in attempting to make customers do without coins amid the shortage. Some stores in the Giant Food chain are reportedly doing the same. And convenience store chain Wawa is reportedly asking shoppers to pay through exact change, debit cards, credit cards or the Wawa mobile app if they can, although the request isn’t mandatory.
Some cash-paying customers might dislike such moves, but putting change on loyalty cards could be a real boon for merchants. For openers, shoppers who don’t already have loyalty cards for a given chain might sign up just to keep from losing their change. Not only will those customers give the grocers some personal information in doing so, but they’ll presumably revisit the retailer in the future to spend their unused change.
Grocers are already tying loyalty programs to the contactless payment systems that many customers seem to favor in a post-pandemic world.
For instance, Kroger launched its Kroger Pay contactless payment option in April. The system allows consumers to link their payment information to their loyalty account at the supermarket chain through an app, then use a QR code for payment at checkouts or self-checkouts.
Similarly, southeast U.S. grocery chain Publix recently unveiled a new loyalty program called Club Publix. It brings together digital features, including a branded digital wallet, the choice to receive eReceipts and early notifications of promotions.
If the U.S. coin shortage grows, more and more retailers could force shoppers to “keep the change” on their loyalty cards. That might be a real positive for retailers.
from News – PYMNTS.com https://www.pymnts.com/news/retail/2020/how-retailers-are-cashing-in-on-a-us-coin-shortage/
JavaScript keeps adding new and neat features. Sometimes, it’s hard to keep up. In this article, I’ll share a couple of cool tips & tricks to keep you up to speed and deepen your JS knowledge.
Imagine having an array with some duplicate items and wanting to filter out only the unique ones.
You could try writing a map or filter to achieve this. Alternatively, ES6 introduces the Set object, which solves this problem in just 1 line of code.
const arrayWithUniqueItems = [...new Set([1, 2, 3, 3,])]
// [1, 2, 3]
Now, this example uses integers, but you can use strings and floating-point numbers as well!
For a little more in-depth knowledge about the Set object, check out this article by Claire-Parker Jones.
Now this is a tricky one.
Shortening your “if” statements can be a great way to simplify your code.
However, if you need to write more complicated statements, you should definitely go for the first option.
// Instead of using this
if (iAmHungry) {
bakeAnEgg()
}
// You can use this
if (iAmHungry) bakeAnEgg()
// Or this
iAmHungry? bakeAnEgg() : 0
Remember, readability & ease-of-use are more important than a couple less lines of code.
A great way of shortening an array is by redefining its length property.
let array = [0, 1, 2, 3, 4, 5, 6, 6, 8, 9]
array.length = 4
// Result: [0, 1, 2, 3]
Important to know though is that this is a destructive way of changing the array. This means you lose all the other values that used to be in the array.
Let’s say you want to combine multiple objects into one object containing them all.
The spread operator ( … ) is a great way to achieve this!
const obj1 = {'a': 1, 'b': 2}
const obj2 = {'c': 3}
const obj3 = {'d': 4}
// Combine them using the spread operator
const objCombined = {...obj1, ...obj2, ...obj3}
// Result: {'a': 1, 'b': 2, 'c': 3, 'd': 4}
Something to keep in mind while using this is that whenever you update one of the objects, it doesn’t reflect those changes in the combined object.
JavaScript can access the current URL using the window.location object. Pretty neat, but even cooler is that this object contains certain parts of the URL as well.
Get access to the protocol/host/pathname/search/and more!
// JavaScript can access the current URL in parts. For this URL:
`https://thatsanegg.com/example/index.html?s=article`
window.location.protocol == `https:`
window.location.host == `thatsanegg.com`
window.location.pathname == `/example/index.html`
window.location.search == `?s=article`
Thanks for reading, look at how much you’ve learned 😄
This article was originally posted on “That’s an Egg” 🍳
5 Must-know Javascript Tips & Tricks was originally published in Prototypr on Medium, where people are continuing the conversation by highlighting and responding to this story.
from Prototypr https://blog.prototypr.io/5-must-know-javascript-tips-tricks-d1a04e9014de?source=rss—-eb297ea1161a—4
The past year has seen remarkable change. As innovation in the field of AI and real-world applications of its constituent technologies such as machine learning, natural language processing, and computer vision continue to grow, so has an understanding of their social impacts.
At our AI-focused Transform 2020 event, taking place July 15-17 entirely online, VentureBeat will recognize and award emergent, compelling, and influential work in AI through our second annual VB AI Innovation Awards.
Drawn both from our daily editorial coverage and the expertise, knowledge, and experience of our nominating committee members, these awards give us a chance to shine a light on the people and companies making an impact in AI.
Our nominating committee includes:
Claire Delaunay, Vice President of Engineering, Nvidia
Claire Delaunay is vice president of engineering at Nvidia, where she is responsible for the Isaac robotics initiative and leads a team to bring Isaac to market for use by roboticists and developers around the world.
Prior to joining Nvidia, Delaunay was the director of engineering at Uber, after it acquired Otto, a startup she cofounded. She was also the robotics program lead at Google and founded two other companies, Botiful and Robotics Valley.
Delaunay has 15 years of experience in robotics and autonomous vehicles leading teams ranging from startups and research labs to Fortune 500 companies. She holds a Master of Science in computer engineering from École Privée des Sciences Informatiques (EPSI).
Asli Celikyilmaz, Principal Researcher, Microsoft Research
Asli Celikyilmaz is a principal researcher at Microsoft Research (MSR) in Redmond, Washington. She is also an affiliate professor at the University of Washington. She received her Ph.D. in information science from the University of Toronto, Canada, and continued her postdoc study in the Computer Science Department at the University of California, Berkeley.
Her research interests are mainly in deep learning and natural language (specifically language generation with long-term coherence), language understanding, language grounding with vision, and building intelligent agents for human-computer interaction. She serves on the editorial boards of Transactions of the ACL (TACL) as area editor and Open Journal of Signal Processing (OJSP) as associate editor. She has received several “best of” awards, including at NAFIPS 2007, Semantic Computing 2009, and CVPR 2019.
The award categories are:
Natural Language Processing/Understanding Innovation
Natural language processing and understanding have only continued to grow in importance, and new advancements, new models, and more use cases continue to emerge.
Business Application Innovation
The field of AI is rife with new ideas and compelling research, developed at a blistering pace, but it’s the practical applications of AI that matter to people right now, whether that’s RPA to reduce human toil, streamlined processes, more intelligent software and services, or other solutions to real-world work and life problems.
Computer Vision Innovation
Computer vision is an exciting subfield of AI that’s at the core of applications like facial recognition, object recognition, event detection, image restoration, and scene reconstruction — and that’s fast becoming an inescapable part of our everyday lives.
AI for Good
This award is for AI technology, the application of AI, or advocacy or activism in the field of AI that protects or improves human lives or operates to fight injustice, improve equality, and better serve humanity.
Startup Spotlight
This award spotlights a startup that holds great promise for making an impact with its AI innovation. Nominees are selected based on their contributions and criteria befitting their category, including technological relevance, funding size, and impact in their sub-field within AI.
As we count down to the awards, we’ll offer editorial profiles of the nominees on VentureBeat’s AI channel The Machine and share them across our social channels. The award ceremony will be held on the evening of July 15 to conclude the first day of Transform 2020.
from Big Data – VentureBeat https://venturebeat.com/2020/07/11/announcing-the-second-annual-venturebeat-ai-innovation-awards-at-transform-2020/
Sorry, we didn’t find anything.
from I need a guide http://inagblog.com/2020/07/jesse-stone/?utm_source=rss&utm_medium=rss&utm_campaign=jesse-stone&utm_source=rss&utm_medium=rss&utm_campaign=jesse-stone
“Our intelligence is what makes us human, and AI is an extension of that quality”. -Yann LeCun, Professor at NYU
Generative AI is a popular topic in the field of Machine Learning and Artificial Intelligence, whose task, as the name suggests, is to generate new data.
There are quite a few ways in which such AI Models are trained , like using Recurrent Neural Networks, Generative Adversarial Networks, Markov Chains etc.
In this article, we are going to look at Markov Chains and understand how they work.We won’t dive deep into the mathematics behind it, as this article is simply meant to get you comfortable with the concept of Markov Chains
Markov Chains are models which describe a sequence of possible events in which probability of the next event occuring depends on the present state the working agent is in.
This may sound confusing, but it’ll become much clearer as we go along in this article. We will be covering the following topics:
A Markov Chain model predicts a sequence of datapoints after a given input data. This generated sequence is a combination of different elements based on the probability of each them occuring immediately after our test data. The length of the input and output data sequences depends on the order of the Markov Chain — which will be explained later in this article.
To explain it simply, lets take an example of a Text Generation AI. This AI can construct sentences if you pass a test word and specify the number of words the sentence must contain.
Before going further, lets first understand how a Markov Chain model for text generation is designed. Suppose you want to make an AI that generates stories in the style of a certain author. You would start by collecting a bunch of stories by this author. Your training code will read this text and form a vocabulary i.e list out the unique words used in the entire text.
After this, a key-value pair is created for each word, where the key is the word itself, and the value is a list of all words that have occured immediately after this key. This entire collection of key-value pairs is basically your Markov Chain model.
Now, lets get on with our example of a Text Generation AI. Here’s a snippet of an example model
This is just a snippet. For the sake of simplicity, I have shown key-value pairs for only 4 words.
Now, you pass it a test word, say “the”. As you can see from the image, the words that have appeared after “the” are “new”, “apple”, “dog”, “cat”,“chair” and “hair”. Since they all have occured exactly once, there is an equal chance of either of them appearing right after “the”.
The code will randomly pick a word from this list. Lets say it picked “apple”. So, now you’ve got a part of a sentence : “the apple”. Now the exact same process will be repeated on the word “apple” to get the next word. Lets say it is “is”.
Now the portion of sentence you have is : “the apple is”. Similarly, this process is run on the word “is” and so on until you get a sentence containing your desired number of words (which is the number of time you will run the program in a loop). Here’s a simplified chart of it all.
As you can see, our output from the test word “the” is “the apple is delicious”. It is also possible that a sentence like “the chair has juice” (assuming “has” is one of the values in the key-value list of the word “chair”) is formed.
The relevance of the generated sentences will directly depend on the amount of data you have used for training. The more data you have, the more vocabulary your model will develop.
One of the major things to note is that the more number of times a particular word occurs after a certain test word in your training data, the higher is the probability of it occuring in your final output.
For example, if in your training data , the phrase “the apple” has occured 100 times, and “the chair” has occured 50 times, in your final output, for the test word “the”, “apple” has a higher probability of occuring than “chair”.
This is based on the basic probability rules
Now, lets look at a term we came across earlier in this section : Order of a Markov Chain
The order of the Markov Chain is basically how much “memory” your model has. For example, in a Text Generation AI, your model could look at ,say,4 words and then predict the next word. This “4” is the “memory” of your model, or the “order of your Markov Chain”.
The design of your Markov Chain model depends on this order. Lets take a look at some snippets from models of different orders
This is the basic concept and working of Markov Chains.
Lets take a look at some ways you can apply Markov Chains for your Generative AI projects
“Talking to yourself afterwards is ‘The Road To Success’. Discussing the Challenges in the room makes you believe in them after a while”
—Generated by TweetMakersAI
Markov Chains are a great way to implement a ML code, as training is quite fast, and not too heavy on an average CPU.
Although you won’t be able to develop complex projects like face generation like that made by NVIDIA, there’s still a lot you can do with Markov Chains in Text Generation.
They work great with text generation as there isn’t much effort required to make the sentences make sense. The thumb rule (as is for most ML algorithms) is that the more relevant data you have, the higher accuracy you will achieve.
Here are a few applications of Text Generation AI with Markov Chains
There are countless things you can do in Text Generation with Markov Chains if you use your imagination.
A fun project that uses Generative AI, is TweetMakers. This site generates fake tweets in the style of certain Twitter users.
As an AI enthusiast, and a meme lover, I believe meme creation is going to be a major application for Generative AI. Check out my blog about the sites which have already started doing so.
Although there’s a lot you can do with Markov Chains , they do have certain limitations. Lets have a look a few of them.
In text generation, Markov Chains can play a huge role. However, there are some minor restrictions to it:
Markov Chains are a basic method for text generation. Although their output can directly be used for various purposes, you will inevitably have to do some post-processing on the output to achieve complex tasks
Markov Chains are a great way to get started with Generative AI, with a lot of potential to accomplish a wide variety of tasks.
Generative AI is a popular topic in ML/AI, so it is a good idea for anyone looking to make a career in this field to get into it, and for absolute beginners, Markov Chains is the way to go.
I hope this article was helpful and you enjoyed it 🙂
Machine Learning Algorithms: Markov Chains was originally published in The Startup on Medium, where people are continuing the conversation by highlighting and responding to this story.
from The Startup – Medium https://medium.com/swlh/machine-learning-algorithms-markov-chains-8e62290bfe12?source=rss—-f5af2b715248—4