If 2023 was all about the rise of generative artificial intelligence (AI) and its entry into mainstream tech conversations, 2024 became the year when AI began displaying its transformative capabilities. What started as a text-based chatbot fad that could respond to users in a human-like fashion, is today powering many major tech products and platforms offering practical use cases. New use cases of the technology were also seen in music and video generation as well as agentic capabilities. And contrary to the opinions of the nay-sayers, the AI bubble did not burst this year.
The year 2024 marked the entry of large language models (LLMs) focused on advanced reasoning, the beginning of the era of AI PCs (Copilot+ PCs if you take Microsoft’s word for it), and accelerated growth of the open-source AI space. However, these are just some of the major events that dominated the headlines this year. Let us take a look at the best and the biggest moments that shaped the AI space in 2024.
OpenAI’s Year of High-Performance AI Models
OpenAI might have started the generative AI trend with its Generative Pre-trained Transformer (GPT) architecture in late 2022, but by the end of 2023, it was clear that the tech giants were not going to stay out of the race for long. Google, Microsoft, Meta, and even Amazon released several AI models, trying to take the crown in benchmark scores.
OpenAI started the year big with its advanced reasoning-focused GPT-4o AI model release in May, which was followed by the GPT-4o Mini in July. The AI firm also ended the year on a high with the launch of the full version of the o1 model and the much-anticipated release of its text-to-video model Sora.
Additionally, the company also introduced its Advanced Voice Mode with Vision to the ChatGPT app, offering newer ways to interact with the chatbot. OpenAI also launched its own search engine dubbed ChatGPT Search, which was integrated within the chatbot platform.
But the biggest coup for the AI firm came in the form of a partnership with Apple, which saw ChatGPT be integrated with Apple Intelligence tools. Following the partnership, OpenAI also released a standalone macOS and Windows app for ChatGPT.
Google’s Diverse Set of AI Offerings
Google also went ballistic with its large number of model releases. In February, the company introduced the Gemini 1.5 series of AI models including the Gemini 1.5 Pro with one trillion parameters. In December, it closed the year by releasing the Gemini 2.0 series, with the Flash model available to everyone in preview, and a larger model reserved for the paid subscribers.
But that was not all the Mountain View-based tech giant did. Google DeepMind, the AI wing of the company, released the Imagen 3 image generation model, and Veo 2 video generation model, and has previewed the music generation AI model MusicLM. Apart from this, the tech giant also released NotebookLM, an AI tool to process large documents that can also create engaging podcasts with two AI hosts.
The company also introduced new features in Gemini. It added a two-way voice communication feature called Gemini Live, integrated the Gemini AI assistant into most of the Google Workspace apps including Gmail, Docs, Slides, and Sheets.
Meta might have been known for its social media platforms before 2024, but this year, the company showcased its capabilities by developing and releasing several small language models (SLMs), many of which were released in open source.
The tech giant introduced several of its Large Language Model Meta AI (Llama) series models including 70B and 30B coding-focused models, the largest open-source model Llama 3.1 405B, as well as multiple instruct models. However, the company’s biggest announcement came with the expansion of its native chatbot Meta AI globally.
Meta AI was added to Facebook’s Messenger, Instagram, and WhatsApp and was expanded to several regions including India in April 2024 before making it globally available by September. The AI-powered chatbot was also added to its Ray-Ban Meta glasses with real-time vision processing capabilities.
Microsoft and the Era of Copilot+ PCs
Even while using AI models of OpenAI, Microsoft was successful in carving an AI niche in the PC space. The Redmond-based tech giant quickly made its desires known when it partnered with Snapdragon (and later with Intel and AMD) to introduce the AI PC classification, which had a mandatory requirement — the addition of a physical Copilot button on the keyboard. Thus came the era of Copilot+ PC, where the company’s native chatbot was integrated into desktops and laptops via the Windows operating system.
Expanding its AI chatbot to millions of users would be considered a success in every business playbook, however, the tech giant was far from done. In 2024, it also integrated Copilot tools into Microsoft 365 products, and added voice and vision capability to the chatbot. Additionally, it also launched the AI-powered Recall feature (in beta) that lets PC users ask the AI questions about past device activity.
Amazon’s Role as AI Aggregator
Many industry analysts had said that Amazon was late to enter the AI space, and while that might be true, the company took a unique route in 2024 to still remain relevant in the AI space. In terms of AI-based releases, the company did not have many standout moments. It did release the Rufus AI tool in the Amazon app that acts as a shopping assistant. It also released the Titan series of AI models and a video generation model for enterprises.
However, the company also quietly took on the role of an aggregator and began integrating AI models from a large number of third-parties to its Amazon Web Services (AWS) platform. It also invested into releasing AI tools that improve the efficiency of the responses and reduce hallucinations. Amazon also bolstered its servers to enable them to handle large volume of AI processing.
Other Notable AI Announcements
While the limelight was on the major AI players in 2024, smaller AI firms did not fail to impress either. Anthropic continued its success with Claude by releasing Claude 3 series early in the year, and the Claude 3.5 series towards the end. The company also launched a desktop app for Mac and Windows in beta, as well as standalone apps for Android and iOS. Additionally, its Tool Use and PDF understanding capabilities made Claude a more capable chatbot in 2024.
Perplexity, the AI-powered search engine, launched a Pro mode that shows detailed responses for complex queries. It also launched a standalone Mac app this year. However, while there were positives, the company’s decision to incorporate ads to even the premium subscribers drew some criticism.
Mistral continued its consistent release of fully open-source AI models even in 2024. It began with the release of the 8x22B Mixture of Experts (MoE) AI models and followed it up with Mixtral Open 2 LLM. The company also surprised developers with the release of Pixtral 12B AI model which comes with computer vision capabilities.
AI in 2025: A Brief Outlook
While we have tried to capture all the major announcements in the AI space in 2024, it is quite impossible to mention every single notable release given the AI fever that is running wild in the tech industry. But now that the year is ending, we expect 2025 to be a similarly action packed year for this technology.
In the coming year, we expect to see the rise of agentic 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 possible and it completing the action without requiring any intervention. That is what AI agents can offer.
Additionally, we also believe the upcoming year will see better implementation of memory function in chatbots, ditching the rudimentary retrieval-augmented generation (RAG) framework. This will lead to chatbots becoming better assistants and companion for users. Real-time video processing might also become more accessible in the coming year. And finally, we believe India will take major strides towards adopting AI in 2025.