Join our daily and weekly newsletters for the latest updates and exclusive content covering cutting-edge AI. Learn more
By any measure, 2024 was the most important year yet for artificial intelligence – at least when it comes to commercializing the technology.
The boom in large language models (LLMs) spurred by the launch of ChatGPT in late 2022 has shown no signs of slowing down, with many new LLMs introduced not only by OpenAI and stalwart tech giants such as Microsoft, Meta and Google , but also by many other startups. and individual developers.
Reports of a slowdown in AI research have proven, if not unfounded, at least certainly exaggerated for now.
Additionally, new technologies beyond the Transformer architecture that underpins most major LLMs have begun to surface, such as Liquid Foundation Templates from Liquid AI.
And eventually, businesses began to fully embrace the “agentic” approach to AI – developing specific AI-powered bots, applications and workflows that can solve specific problems independently or with less human management than the typical LLM back and forth. chatbots.
Narrowing down the year’s news stories to a top 14, let alone a top 10 or top 4, was a frustrating effort. But I went ahead and tried, cheating slightly by combining several stories into larger themes. In my eyes, here is what will have the biggest impact from this year:
1. OpenAI has expanded far beyond ChatGPT
The company, arguably most responsible for ushering in the AI generation era, hasn’t missed a beat this year, despite intensifying competition from upstarts and traditional tech alike, including from its own investor and Microsoft partner.
Model o1: OpenAI has released its first new family of large general-purpose models beyond its GPT series, the o1 “reasoning” serieswhich allocates more time to processing complex prompts, resulting in greater accuracy. It is particularly effective in scientific, coding, and reasoning tasks.
Model o3: He followed the September o1 template with a blockbuster year-end ad of the same more advanced o3 model. While this won’t be available to the public or even third parties until early 2025, it shows that OpenAI isn’t resting on its laurels.
Search ChatGPT: This feature, initially launched as an invite-only standalone product called SearchGPT before being bundled into ChatGPT proper, enables real-time web information retrieval in ChatGPT and refined presentation of search results, improving its usefulness for up-to-date queries and taking on Google, Bing and newcomer Perplexity.
Cloth: Introduced in October, Cloth extends the ChatGPT interface beyond that of a conversational interface to a desktop-style pane that can dynamically update content upon user request, such as editing a document or editing ‘a coding project. Of course, it was difficult not to see this as a reaction, or at least a comparable characteristic, to The Anthropic artifacts announced several months ago.
Sora: After almost a year of teasing us with its well-guarded video generator model, At the beginning of December, OpenAI finally launched Sora to the general publicquickly inviting a wide range of reactions as it sought to differentiate itself in a highly competitive AI video space with a unique and well-thought-out interface and storyboard functionality.
2. Open source AI has taken off
Llama 3 and 3.1: Meta introduced Llama 3 in Aprilsetting a new standard of performance in open source AI, then quickly followed with Llama 3.1 in July with 405 billion parameters. Llama 3.1 releases were used to power Meta AI, the company’s built-in assistant on platforms such as WhatsApp, Messenger, Instagram and Facebook, with the aim of becoming the most widely used AI assistant.
Llama 3.3: Released in December 2024, Llama 3.3 delivers performance comparable to larger models, but at a fraction of the compute cost, making it more accessible to enterprise applications.
Meanwhile, Chinese models like Alibaba’s Qwen-2.5 family And The new V2.5 of DeepSeek And R1-Lite Overview appeared seemingly out of nowhere at the top of some of the benchmark rankings, and Nvidia itself has gone beyond providing graphics cards and software architectures to release its own powerful open source software. Nemotron-70B model.
Nous Research, a small structure from San Francisco that wants to offer more personalized and less restrictive AI models as open source, also launched several cool new ideas.
And let’s not forget France Mistralwhich has rapidly expanded its own open source and proprietary AI offerings.
3. Google’s Gemini series has become a strong contender for the best available
In the comeback story of the year, Google’s Gemini series of AI models, which were once mocked for their weird image generations and criticized for being too “woke”, have come back with a vengeance with new versions more powerful that now top third-party performance rankings. and are increasingly attractive to developers and businesses.
Google presented Gemini 2.0 Flasha multi-modal AI model that supports streaming video analytics and can see and report what you’re doing on your screen, and track it with Thought Flash Gemini 2.0 which rivals OpenAI’s o1 and o3 reasoning models.
4. Agentic AI has swept the enterprise
As the year progressed, “agentic” AI went from being a buzzword to becoming a veritable series of product announcements and major initiatives from major enterprise software vendors. Take for example:
Salesforce Agentforce 2.0: Salesforce unveiled Agentforce 2.0 a few days ago, an advanced AI agent program to improve the reasoning, integration and personalization functionalities of its CRM and commercial offers, as well as Softsignificantly improving the company’s productivity tools.
Joule of SAP: SAP converted its Joule chatbot into a AI agent powered by open source extended language models (LLMs)), promoting innovation and efficiency in businesses.
Google’s Astra Project: As part of the Gemini 2.0 initiative, Google launched Project Astra, an AI assistant designed to provide real-time contextual responses by leveraging Google’s suite of services, aimed at improving productivity and user decision-making.
My big prediction for 2025: AI-generated content will reign supreme
Building on these advances, 2025 is poised to see the proliferation of AI-generated content in the business and consumer arenas, especially as everyone from OpenAI to Meta, Google, Microsoft, Apple and even Elon Musk’s xAI now has AI image generators integrated into their offers.
This expansion will streamline content creation, improve personalization, and improve efficiency across various industries.
Additionally, we anticipate the first large-scale deployments of large language models (LLM) and AI-driven generative robotics in commercial and consumer environments, revolutionizing automation and human-robot interactions.
That’s all in the latest #AIBeat newsletter for 2024. Thanks for reading, writing, subscribing, sharing, commenting and being here with us. Looking forward to sharing more and hearing more from you all in 2025.
Happy Holidays and Happy New Year from all of us at VentureBeat to you and your loved ones.
