GOOGLE

YouTube Algorithm: 6 Questions Answered

Published

3 years ago

November 30, 2020

youtube algorithm 6 questions answered via mattgsouthern

YouTube is sharing more details about how its search and recommendation algorithms work in a new video where the company answers questions from users.

The YouTube team published a similar video earlier this month, though its newest video answers an all-new set of questions.

There’s quite a bit of material to go over so let’s get right into it.

Impact of Changing Titles & Thumbnails

If a video isn’t performing well, would it help to change the title and thumbnail? Or would that make the algorithm lose confidence in the video?

YouTube absolutely recommends changing the way a title or thumbnail looks, as it can be an effective way to get more views.

That’s generally because the video looks different to viewers and that’s going to change how people interact with it when it’s offered in their recommendations. YouTube’s algorithm then responds to the change in user behavior, not the act of changing the title or thumbnail.

The act of changing a title or thumbnail does not inherently trigger YouTube to increase the impressions for a video. It’s all about how users respond to the change.

In general, making changes to a video is only recommended when it has both a lower click-through rate and it’s receiving fewer views and impressions than usual.

Algorithm Response to Old/Inactive Subscribers

Can old/inactive subscribers negatively affect the performance of a video? The concern is this could lead to a lower CTR, which may result in the video not being recommended as much.

YouTube’s recommendation algorithm doesn’t focus on the subscription feed as a primary signal. The algorithm is focused on how well a video performs in the context it’s shown in.

Ranking on the home page, for example, is based on how well that video performed when shown on other users’ home pages.

YouTube’s algorithm understands which viewers have not watched a channel’s content in a long time, and will avoid showing content from that channel to inactive subscribers.

So inactive subscribers are not something channel owners should be worried about.

How is a total subscriber count relevant if YouTube won’t push out content to all subscribers based on their inactivity/lack of engagement on the channel. Shouldn’t videos be pushed out to someone unless they unsubscribe?

YouTube’s recommendation system does not push videos out to anyone. What it does is pull videos in and ranks them for users based on what they’re most likely to watch.

Subscribers are one of many signals used to rank videos for users. It testing, YouTube found prioritizing content from channels a user subscribes to dramatically reduces how many videos users watch and how often they come back to YouTube.

That’s why YouTube’s recommendation algorithm is designed to recommend content users are likely to watch, regardless of whether it’s published by channels a user subscribes to.

YouTube Search Results

How does YouTube rank search results?

Just like Google’s search engine, search on YouTube has a similar goal where it wants to show users the most relevant results for their queries.

Videos are ranked in YouTube search according to a variety of factors, but the most important factors are relevance and performance.

Relevance is how well the title, description, and content of a video match the user’s query.

Performance is related to which videos users chose to watch after conducting similar queries.

YouTube’s algorithm also considers engagement metrics such as how long and how much of a video users choose to watch.

To clarify, YouTube’s search results are not a list of the most viewed results for a given query. It’s more about which videos are the most relevant and which videos a user is most likely to watch.

Multiple Languages on the Same Channel

Can uploading videos in two different languages on the same channel affect how videos from that channel are recommended by YouTube?

Uploading in different languages on the same channel can be confusing to viewers. For that reason, YouTube recommends creating separate channels for each language.

However, if the channel specifically caters to an audience that speaks multiple languages, then keeping all content on the same channel makes sense.

Importance of Watch-Time

Does it take a certain amount of hours of watch-time before a video is recommended by YouTube’s algorithm?

There’s no particular threshold a video needs to meet before it starts getting recommended.

Channels may notice some of their videos gaining momentum months after being published because it’s common for users to show interest in old videos. This could be because a particular topic is rising in popularity, or new viewers of a channel may be going back and watching previous videos.

Most users do not watch videos in the order of most recent, or decide what they want to watch based on when it was published. So a user’s home page will often contain videos published weeks, months, or even years ago.

See YouTube’s full Q&A video below:

Searchenginejournal

GOOGLE

Google’s Next-Gen AI Chatbot, Gemini, Faces Delays: What to Expect When It Finally Launches

Published

5 days ago

December 4, 2023

Max

In an unexpected turn of events, Google has chosen to postpone the much-anticipated debut of its revolutionary generative AI model, Gemini. Initially poised to make waves this week, the unveiling has now been rescheduled for early next year, specifically in January.

Gemini is set to redefine the landscape of conversational AI, representing Google’s most potent endeavor in this domain to date. Positioned as a multimodal AI chatbot, Gemini boasts the capability to process diverse data types. This includes a unique proficiency in comprehending and generating text, images, and various content formats, even going so far as to create an entire website based on a combination of sketches and written descriptions.

Originally, Google had planned an elaborate series of launch events spanning California, New York, and Washington. Regrettably, these events have been canceled due to concerns about Gemini’s responsiveness to non-English prompts. According to anonymous sources cited by The Information, Google’s Chief Executive, Sundar Pichai, personally decided to postpone the launch, acknowledging the importance of global support as a key feature of Gemini’s capabilities.

Gemini is expected to surpass the renowned ChatGPT, powered by OpenAI’s GPT-4 model, and preliminary private tests have shown promising results. Fueled by significantly enhanced computing power, Gemini has outperformed GPT-4, particularly in FLOPS (Floating Point Operations Per Second), owing to its access to a multitude of high-end AI accelerators through the Google Cloud platform.

SemiAnalysis, a research firm affiliated with Substack Inc., expressed in an August blog post that Gemini appears poised to “blow OpenAI’s model out of the water.” The extensive compute power at Google’s disposal has evidently contributed to Gemini’s superior performance.

Google’s Vice President and Manager of Bard and Google Assistant, Sissie Hsiao, offered insights into Gemini’s capabilities, citing examples like generating novel images in response to specific requests, such as illustrating the steps to ice a three-layer cake.

While Google’s current generative AI offering, Bard, has showcased noteworthy accomplishments, it has struggled to achieve the same level of consumer awareness as ChatGPT. Gemini, with its unparalleled capabilities, is expected to be a game-changer, demonstrating impressive multimodal functionalities never seen before.

During the initial announcement at Google’s I/O developer conference in May, the company emphasized Gemini’s multimodal prowess and its developer-friendly nature. An application programming interface (API) is under development, allowing developers to seamlessly integrate Gemini into third-party applications.

As the world awaits the delayed unveiling of Gemini, the stakes are high, with Google aiming to revolutionize the AI landscape and solidify its position as a leader in generative artificial intelligence. The postponed launch only adds to the anticipation surrounding Gemini’s eventual debut in the coming year.

GOOGLE

Google Brings Bard Students Math and Coding Education in the Summer

Published

6 months ago

June 8, 2023

Entireweb News Bot

$Google Brings Bard Students Math and Coding Education in the Summer$

Google is stepping up its AI efforts this summer by sending Bard, its high-profile chatbot, to summer school. The aim? To boost the bot’s math and coding smarts. These developments are excellent news— when Bard first debuted, it was admittedly not a finished product. But Google is steadily plugging away at it, and have now implemented implicit code execution for logical prompts, and handy Google Sheets’ integration to take it to the next level.

Thanks to implicit code execution, Bard can respond to inquiries requiring calculation or computation with Python code snippets running in the background. What’s even more amazing is that coders can take this generated code and modify it for their projects. Though Google is still apprehensive about guaranteeing the accuracy of Bard’s answers, this feature is said to improve the accuracy of math and word problems by an impressive 30%.

In addition to this, Bard can now export directly to Sheets when asked about tables. So, you don’t need to worry about copying and pasting, which comes with the risk of losing formatting or data.

From the company’s I/O keynote address, it is clear that they are focused on making the most of what Bard can offer. As they continue to speak highly of the chatbot, we’re sure to expect more features and capabilities when the summer comes around.

GOOGLE

Google Bard vs. ChatGPT: which is the better AI chatbot?

Published

10 months ago

February 20, 2023

Entireweb News Bot

Google Bard vs. ChatGPT: which is the better AI chatbot?

Google Bard and ChatGPT are two of the most prominent artificial intelligence (AI) chatbots available in 2023. But which is better? Both offer natural language responses to natural language inputs, using machine learning and millions of data points to craft useful, informative responses. Most of the time. These AI tools aren’t perfect yet, but they point to an exciting future of AI assistant search and learning tools that will make information all the more readily available.

As similar as these chatbots are, they also have some distinct differences. Here’s how ChatGPT and Google Bard measure up against one another.

Which is better, Google Bard or ChatGPT?

This is a tricky question to answer, as at the time of writing, you can only use Google Bard if you’re part of a select group of early beta testers. As for its competition, you can use ChatGPT right now, completely for free. You may have to contend with a waitlist, but if you want to skip that, there’s a paid-for Plus version offering those interested in a more complete tool the option of paying for the privilege.

Still, when Google Bard becomes more widely available, it should offer credible competition for ChatGPT. Both use natural language models — Google Bard uses Google’s internal LaMDA (Language Model for Dialogue Applications), whereas ChatGPT uses an older GPT-3 language model. Google Bard bases its responses to questions on more recent data, with ChatGPT mainly trained on data that was available prior to 2021. This is similar to how Microsoft’s Bing Chat works.

We’ll have to reserve judgment on which is the more capable AI chatbot until we get time to play with Google Bard ourselves, but it looks set to be a close contest when it is more readily available.

Are Google Bard and ChatGPT available yet?

As mentioned, ChatGPT is available in free and paid-for tiers. You might have to sit in a queue for the free version for a while, but anyone can play around with its capabilities.

Google Bard is currently only available to limited beta testers and is not available to the wider public.

Banner of Google Bard intro from February 6.

What’s the difference between Google Bard and ChatGPT?

ChatGPT and Google Bard are very similar natural language AI chatbots, but they have some differences, and are designed to be used in slightly different ways — at least for now. ChatGPT has been used for answering direct questions with direct answers, mostly correctly, but it’s caused a lot of consternation among white collar workers, like writers, SEO advisors, and copy editors, since it has also demonstrated an impressive ability to write creatively — even if it has faced a few problems with accuracy and plagiarism.

Still, Microsoft has integrated ChatGPT into its Bing search engine to give users the ability to ask direct questions of the search engine, rather than searching for terms of keywords to find the best results. It has also built it into its Teams communications tool, and it’s coming to the Edge browser in a limited form. The Opera browser has also pledged to integrate ChatGPT in the future.

ChatGPT	Google Bard
Accessible through ChatGPT site. Only text responses are returned via queries.	Integrated with Google Search. You only need to change a Google setting to get your regular search results when using Google Bard AI, and vice versa.
ChatGPT produces answers from its trained database from 2021 and before.	Google Apprentice Bard AI will be able to answer real-time questions.
Based on GPT (Generative Pre-trained Transformer).	Based on LaMDA (Language Model for Dialogue Applications).
Service has a free and paid plan option (called ChatGPT Plus).	Service is free.
Has built-in plagiarism tool called GPT-2 Output Detector.	No built-in plagiarism detection tool.
Available now	Still in beta test phase

Google Bard was mainly designed around augmenting Google’s own search tool, however it is also destined to become an automated support tool for businesses without the funds to pay for human support teams. It will be offered to customers through a trained AI responder. It is likely to be integrated into the Chrome browser and its Chromium derivatives before long. Google is also expected to open up Google Bard to third-party developers in the future.

Under the hood, Google Bard uses Google’s LaMDA language model, while ChatGPT uses its own GPT3 model. ChatGPT is based on slightly older data, restricted in its current GPT3 model to data collected prior to 2022, while Google Bard is built on data provided on recent years too. However, that doesn’t necessarily make it more accurate, as Google Bard has faced problems with incorrect answers to questions, even in its initial unveiling.

ChatGPT also has a built-in plagiarism checker, while Google Bard does not, but Google Bard doesn’t have the creative applications of ChatGPT just yet.