Social Throwback 2023: Tech giants integrate AI tools to enhance advertiser and user experience

When ChatGPT was making the rounds last year, Microsoft soon announced a multi-year, multi-billion dollar investment in OpenAI in January 2023. This deal marked the third phase of the partnership between the two companies, following the tech giant's previous investments in 2019 and 2021. The company said the renewed partnership would accelerate breakthroughs in AI and help both companies commercialize advanced technologies in the future. 

This set off a tiff between other giants like Google, Twitter (X), and Meta, with each trying to enhance user experience through AI-driven integrations. 

Alphabet Inc.’s Google, which reoriented the company around AI six years ago to organize the world’s information, has invested in AI across the board, with Google AI and DeepMind. According to the report, 80% of Google’s customers are already using AI-powered Search ad products.

We uncover all the AI developments made by social media and tech giants in the year 2023.

Google: Bard, integration in Workspaces & more

In February 2023, the company introduced an experimental conversational AI service called Bard to compete with Microsoft-backed AI chatbot ChatGPT. CEO Sundar Pichai, in a blog post, mentioned that it could be an ‘outlet for creativity and curiosity’.

Since then, it has expanded its availability to new regions, including Brazil and Europe, and now supports prompts in 40 additional Indian regional languages. The updates enhance user interaction by allowing them to listen to Bard's responses, customize the tone and style of responses, pin and rename conversations, share their chat with others through shareable links, and incorporate images in prompts using Google Lens capabilities. 

Recently, it was revealed that it is set to be powered by Gemini, the company’s advanced AI model, giving the chatbot more advanced reasoning, planning, understanding, and other capabilities. The AI model will be run on anything ranging from mobile devices to data centers and will be rolled out to Bard over two phases. 

It is equipped to respond to inquiries regarding YouTube videos. Initially, with the capability to analyze YouTube videos, it can now furnish specific answers to queries on a video's content with the introduction of the YouTube Extension in September.

Google also extended its generative AI search feature, the Search Generative Experience (SGE), to U.S. teenagers aged 13-17 through Search Labs. The SGE introduced a conversational mode for more natural interactions. Google launched a feature to provide context to search results and enhance its AI to better identify false or offensive queries. This initiative allows teens to access the AI-powered search experience via the Google app or Chrome desktop platform, aiming to improve the search experience for younger users.

It also introduced AI-powered writing features in its Workspace, enabling users to generate text directly in Docs and Gmail, as well as across other apps like Slides, Sheets, Meet, and Chat for drafting, proofreading, data analysis, and streamlining workflows. It enhanced its Maps service, featuring several AI-driven features aimed at simplifying user searches, exploration of their surroundings, and navigation to their destinations. 

In November 2023, it introduced AI tools for advertisers in the US. Through the AI-driven Product Studio, merchants and marketers can harness the power of text-to-image AI technology to generate product visuals effortlessly by inputting their desired image descriptions.

YouTube: Advertising solutions, Music Incubator & more

In October, YouTube launched ‘Spotlight Moments,’ an advertising solution utilizing Google AI to target ads within trending videos associated with cultural events. After this, advertisers could showcase their ads on a dedicated YouTube channel with updated playlists. Media company GroupM is the first to provide clients access to this AI-driven feature, part of YouTube's broader initiatives like Video Reach and Video View, strategically using Google AI for more effective and relevant advertising.

It also announced that users can make their own AI playlist art. Users can click on the default picture on a playlist to choose from themes such as travel, fantasy, animals, and more, followed by a list of prompts. 

It launched the Music AI Incubator, working with music labels like Universal Music Group to establish policies for AI-generated content, focusing on how such music is treated on the platform and monetization opportunities for artists and companies. 

It unveiled new AI-powered features for content creators, including ‘Dream Screen’, which allows AI-generated backgrounds for Shorts, while 'Aloud' enables multilingual dubbing. YouTube Create offers editing tools, 'Creator Music' assists in music searches, and 'YouTube Studio' uses AI for personalized video ideas, simplifying content creation.

The platform has been testing early versions of AI-powered tools in YouTube Studio with creators, and it has been suggested that more than 70% of those surveyed have said it’s helped them develop and test ideas for videos.

Microsoft: Bing search engine, Image Creator

After Google’s Bard, Microsoft announced that it would launch a new version of Bing, with ChatGPT-like AI powers. The company estimated that there are 10 billion search queries a day, and half of them go unanswered. The Bing search engine and Edge browser aimed to deliver better search results, more complete answers, a new chat experience, and the ability to generate content. 

Following this, it launched Bing Image Creator, new AI-powered visual Stories, and updated Knowledge Cards to the new Bing and Edge preview. Powered by an advanced version of the DALL∙E model from their partners at OpenAI, Bing Image Creator allows users to create an image by using their own words to describe the picture they want to see. 

It has been adding features to SwiftKey by tying it with Microsoft Bing, where users can make AI Stickers with Image Creator using their selfies and photos and send it to friends and family on WhatsApp and Messenger.

In November, the company rebranded its AI-powered chatbot, Bing Chat, to Copilot in Bing. The premium version targeted at corporate users, previously named Bing Chat Enterprise, has also undergone a name change to Copilot.

Amazon: Image generation for advertisers, chatbot, modifications to Alexa

In September, Amazon announced the launch of generative AI to Alexa to inculcate the whole experience with human-like interactions. The AI model aimed to power improved experiences across its Echo family of devices and provide a more nuanced version.

The model can interact by focusing on body language, gestures, and eye contact. It can combine multiple requests into one interaction and can also carry on conversations without using the word 'Alexa'.

In October, the company started testing an AI-powered image generation tool designed for advertisers. The tool enables advertisers to produce backgrounds based on product descriptions and themes. It is touted to be beneficial for advertisers of all scales, offering an easy-to-use image generation feature that doesn't require technical expertise.

Named the ‘Titan Image Generator’, it was officially rolled out in November and can generate fresh images based on textual descriptions or modify existing images.

Alongside the image generator, Amazon’s cloud business AWS launched a chat tool, Amazon Q, that will help businesses ask questions specific to their companies. The feature will help users get answers to their business using their data and will be able to access Amazon Q with AWS Management Console or individual companies’ documentation pages, developer environments like Slack, and other third-party apps.

Meta: Advertiser-centric tools, translator & more across family of apps

Meta built a multi-stage AI review system to classify content for advertisers in Facebook and Instagram Feeds to ensure brand suitability controls are in place. The classification is not only text but also video and images to determine if it meets monetization policies.

Having invested in AI for over a decade, the company has worked on an ad creation tool this year that could save spending on the creation and also provide variety instead of using a single image for a campaign. Launched in October, the new features, including Background Generation, Image Expansion, and Text Variations, can add to the AI-powered experiences and tools that Meta is building for businesses.

With this, it has introduced Meta Lattice, a new AI model to predict an ad’s performance across a variety of datasets and optimization goals. Its aim is to improve performance across all ad types and ad surfaces, align with advertiser objectives, and utilize the rapid expansion of high-growth areas, like short-form videos, to provide enjoyable experiences to people.

Its AI model, SeamlessM4T can translate up to 100 languages, and the company aims to integrate these developments into Facebook, Instagram, WhatsApp, Messenger, and Threads. It also worked on introducing a generative AI chatbot across its family of apps named Meta AI on WhatsApp, Messenger, and Instagram direct messages. This AI assistant answers questions with real-time information, generates photo-realistic images, and more. 

Meta also launched Emu Video, allowing users to curate a four-second animated clip based on a caption, image, or photo paired with a description.

The company rolled out an AI stickers feature on WhatsApp to a limited group of beta testers in August. Users will be prompted to enter a description that will be used to generate the sticker. They will also be presented with a set of AI-generated stickers based on the descriptions they had previously entered. 

The platform might assist users in the future using secure service from Meta to provide more relevant and helpful responses with the help of a new feature that will answer queries from users using AI-generated messages. It aims to provide more personalised and effective solutions to user queries.

Snapchat: My AI chatbot, Dreams

Snapchat rolled out My AI in April 2023, a chatbot built with OpenAI’s GPT technology, and provided it to Snapchat+ subscribers. In a chat conversation, My AI can answer trivia questions and offer advice and suggestions. The chatbot has also received some backlash due to its controversial responses. Following this, the platform updated some of its safety enhancements and learnings.

It also experimented with new ways that My AI can surface useful information at the right moment during conversations, including early testing of sponsored links to connect the community and partners with indicated potential interest in their offerings. This led to the platform partnering with Microsoft to incorporate ads in the chatbot. If a user asks the bot for dinner suggestions, it might respond with a sponsored link from a local restaurant or a food delivery app.

Snap Research developed a model called SnapFusion that shortens the runtime from text input to image generation on mobile to under two seconds. Moreover, it launched 'Dreams,' a feature allowing users to enhance selfies by adding AI-generated backgrounds for ‘fantastical’ scenarios, in August. Users can attach or take selfies and use AI to incorporate imaginative backgrounds, expanding on its existing filter feature. 

Twitter (X), LinkedIn, Quora & Pinterest

In November 2023, Twitter (X) owner Elon Musk’s AI startup, xAI, seemed to be developing its variant of ChatGPT. Musk teased xAI's AI system, Grok, which is designed to answer questions conversationally. X promises that Grok will have ‘personality’ and answer questions, that other AI chatbots shy away from and it will be able to answer with up-to-date information on certain topics. The beta version is now rolled out to all Premium+ subscribers in the US.

On the other hand, Quora’s Poe, its AI chatbot, rolled out a revenue system for creators building AI bots on the platform. CEO Adam D'Angelo stated that there are two major ways to get revenue; the first is getting people to subscribe to Poe, and the platform will share a portion of revenue immediately. The second way will be rolled out shortly, which involves the bot creator allotting a designated fee per message, and Quora will pay the message fee.

To become more inclusive, Pinterest rolled out AI body type technology that can identify various body types among the billions of images across its platform’s search results. The technology uses shape, size, and form to identify various body types, influencing the way its algorithms enhance inclusivity in women's fashion and wedding-related content across related feeds and search results. It is also testing a new consumer tool that will let users filter searches based on different body types. The tool will work on women’s fashion and wedding ideas at launch.

According to a LinkedIn-commissioned survey, 55% of marketers plan to use generative AI to increase their efficiency so they can focus on higher-value work. The platform launched an AI-generated copy generator for campaigns. The Mircosoft-backed company uses advanced OpenAI GPT models to leverage data from users' LinkedIn Page and Campaign Manager settings, like objective, targeting criteria, and audience, to suggest ad headlines and copy to help jumpstart the campaigns.

The professional network also developed an AI-based tool that helps users write their posts. 

From AI-powered chatbots to advertising solutions aimed at personalizing content creation and user experiences, tech companies like Google, Microsoft, Meta, and more will be attempting to grab users' attention in 2024.