If 2023 is all about the rise of generative artificial intelligence (AI) and its entry into mainstream technology dialogue, 2024 will be the year when AI begins to demonstrate its ability to change. Originally a text-based chatbot fashion that can react to users in a human way, it is today powering many major technology products and platforms, providing practical use cases. New use cases for the technology are also seen in music and video generation as well as in proxy functionality. Contrary to the opinions of violators, this year’s AI bubble has not burst.
2024 marks the entry of large language models (LLMS), focusing on advanced reasoning, the beginning of the AI PC era (if you say about Microsoft’s word for this, Copilot+ PC), and the growth of the open source AI space has accelerated. But these are just some of the big events in the headlines this year. Let’s take a look at the best and biggest moments to shape the AI space in 2024.
Openai’s high-performance AI model
Openai may start the generated AI trend with its generated pre-trained transformer (GPT) building at the end of 2022, but by the end of 2023, it is clear that the tech giants will not stay away from the competition for a long time. Google, Microsoft, Meta and even Amazon have released several AI models in an attempt to get the crown with benchmark scores.
Openai was released in May with its advanced reasoning-focused GPT-4O AI model followed by the July GPT-4O Mini. AI also ended the year with the release of the video model Sora from the full year of O1 models and the highly anticipated text.
Additionally, the company has introduced advanced voice modes with vision to the Chatgpt application and provides new ways to interact with chatbots. Openai also launched its own search engine, called ChatGpt Search, which is integrated into the chatbot platform.
But the biggest coup for AI companies was the form of a partnership with Apple, which integrated Chatgpt with Apple Intelligence tools. Following the partnership, OpenAI also released a standalone MacOS and Windows application for Chatgpt.
Google’s diverse AI products
Google also released ballistics through a large number of models. In February, the company launched the Gemini 1.5 series AI models, including the Gemini 1.5 Pro, which includes a trillion parameters. In December, it ended the year with the release of the Gemini 2.0 series, with everyone in the preview able to use the Flash model and retaining a larger model for paid subscribers.
But this is not all the tech giants based on mountain views. The company’s AI wing Google DeepMind released the Imagen 3 image generation model and the VEO 2 video generation model, and previewed the music-generation AI Model Musiclm. Apart from that, the tech giant has released NotebookLM, an AI tool for handling large documents and can also create engaging podcasts using two AI hosts.
The company has also introduced new features in Gemini. It adds a two-way voice communication feature called Gemini Live, integrating Gemini AI assistant into most Google Workspace applications, including Gmail, Docs, Slide, and Sheets.
META may be known for its social media platform until 2024, but this year the company demonstrates its capabilities by developing and releasing multiple small language models (SLMs), many of which have been released in open source.
The tech giant has launched its series of models, Meta AI (Llama) in several large language models, including 70B and 30B coding-centric models, the largest open source model, Llama 3.1 405B, and multiple indicator models. However, the company’s biggest announcement is to expand its local chatbot Meta AI globally.
Meta AI was added to Facebook’s Messenger, Instagram and WhatsApp and expanded to multiple regions including India in April 2024 before being acquired worldwide in September. AI-powered chatbots are also added to ray element glasses with real-time visual processing.
The era of Microsoft and Copilot+ PC
Even using OpenAI’s AI model, Microsoft has successfully carved the AI mines in the PC space. The Redmond-based tech giant quickly knew its desire when it partnered with Snapdragon (and later with Intel and AMD) when introducing the AI PC classification, which had mandatory requirements – adding a physical copy to the keyboard button. So the era of Copilot+ PC is the era of the company’s local chatbots integrating into desktops and laptops through the Windows operating system.
Scaling its AI chatbot to millions of users will be considered a success for every commercial script, but the tech giant is far from complete. In 2024, it also integrates the Copilot tool into Microsoft 365 products and adds voice and visual features to the chatbot. Additionally, it launches an AI-driven recall feature (in Beta), which allows PC users to ask questions about AI about past device activity.
Amazon’s role as an AI aggregator
Many industry analysts have said Amazon has entered the AI space late, and while this may be true, the company has taken a unique route in 2024 and remains relevant in the AI space. As far as AI-based distributions are concerned, the company doesn’t have many great moments. It does release the Rufus AI tool in Amazon apps, which is for Shopping Assistant. It also released the Titan series AI models and enterprise video generation models.
However, the company also quietly took on the role of aggregator and began combining AI models from a large number of third-party to its Amazon Web Services (AWS) platform. It also invests in unleashing AI tools to improve responsiveness and reduce hallucinations. Amazon also supports its servers, allowing it to handle a lot of AI processing.
Other famous AI announcements
Although the focus is on the main AI players in 2024, the smaller AI companies are not impressed either. Anthropic released the Claude 3 series early this year through Claude and released the Claude 3.5 series on Claude. The company has also launched a desktop app for Beta for Mac and Windows, as well as standalone apps for Android and iOS. In addition, its tool usage and PDF understanding capabilities make Claude a more capable chatbot in 2024.
The confusion of AI-powered search engines initiates Pro mode, which displays detailed responses to complex queries. It also launched a standalone Mac app this year. But despite the positive attitude, the company’s decision to include ads in even premium subscribers has attracted some criticism.
Even in 2024, Mistral continues its fully open source AI model. It begins with the release of the 8x22b mixture of the Expert (MOE) AI model and then started using the Mixtral Open 2 LLM. The company also surprised developers with the release of the PixTral 12B AI model that released computer vision capabilities.
AI in 2025: A brief prospect
While we are trying to capture all the major announcements in the AI space in 2024, every famous version is unlikely to mention, given the wild AI fever in the technology industry. But now that the year is over, we expect 2025 to be a year of similar action for this technology.
In the coming year, we hope to see the rise of proxy AI and its integration into platforms and devices. Imagine asking your chatbot to book a movie ticket or buy a product at the lowest price and completing the lawsuit without any intervention. This is what AI agents can offer.
In addition, we also believe that the upcoming year will better implement memory capabilities in chatbots, thus abandoning the Generation Generation Generation (RAG) framework of basic retrieval. This will lead to chatbots becoming better assistants and companions for users. Real-time video processing may also be easier to access in the coming year. Finally, we believe India will make great progress in adopting AI in 2025.
