2023 Timeline of AI progress, AI Magazine highlights some of the leading, newly released AI innovations – and the companies that developed them. There has been no shortage of new Artificial Intelligence (AI) innovations this year (2023), with plenty more to come in 2024 & beyond.
New AI: 10. MicroSoft Copilot, 9. Claude 2, 8. GPT-4, 7. Bard, 6. IBM Watsonx, 5. Llama 2, 4. Duet AI, 3. Adobe Firefly 2, 2. DALL-E 3, 1. Gemini AI.
Year in Review. It has been a very busy year for the enterprises developing, deploying and harnessing AI, from new alliances to new – and old – leadership, the sector has experienced plenty of transformation in such a short space of time. Indeed, even world leaders are more attuned to the conversation and are exploring ways in which they can develop AI more safely in collaboration with technology businesses.
With this in mind, here are some of the most talked-about AI innovations from 2023. This list is ranked in descending order, according to launch date.
Launched early in 2023, Microsoft Copilot really adopts the ‘AI-as-copilot’ ethos as its AI system aims to use AI to transform global businesses via efficiency and ease of use. Copilot is designed to uniquely incorporate the context and intelligence of the web, an individual’s work data to provide better assistance – with privacy and security at the forefront. It is available in Windows 11, Microsoft 365 and in web browsers via Edge.
New research from Microsoft has gone on to explore the 4 types of ‘digital humans’ that we can expect to interact with in the near future.
9. Claude 2 AI Assist
Developed by Anthropic, Claude 2 is a next-generation AI assistant for individual and business tasks, regardless of scale. The company describes the AI model as a “friendly, enthusiastic colleague” or “personal assistant” that can be instructed in natural language to help with a wide range of tasks.
The company’s latest model has improved performance, longer responses and can be accessed via Apps Programming Interface (API) as well as a new public-facing beta website, Claude.ai. It is also capable of tasks from sophisticated dialogue and creative content generation to detailed instruction.
8. GPT-4 Chat GPT
Very much anticipated after the successful launch of OpenAI’s first ChatGPT model, a faster GPT-4 model was released in 2023. Upon release, the company stated that it exhibits human-level performance on various professional and academic benchmarks from the US bar exam to SAT school exams.
The widespread interest in Large Language Models (LLMs) like GPT-4 has created huge gains in interest around Neuro-Linguistic Programming (NLP) and deep learning, influencing other companies to want to invest in AI, GPT-4 surpasses GPT-3.5 by a significant margin and exhibits an increased ability to handle complex tasks and more nuance-type instructions.
7. Bard Chat Service
Bard is an experimental, conversational AI tool by Google. Designed to be a Chat service, its purpose is to combine the world’s knowledge with power, intelligence and creativity with LLMs. Bard draws on information from the web to provide high-quality responses, in addition to being used as a creative outlet that helps to explain new discoveries.
Despite a competitive landscape, a Reuters report indicates that Bard’s web traffic increased by 2% in October 2023, reaching 8.7 million, signalling a shift in how consumers gather information.
6. IBM watsonx Gen AI Tools
IBM Watsonx.ai is a studio of integrated tools for working with generative AI capabilities that are powered by foundation models and for building machine learning models. Its data platform includes three core components and a set of AI assistants designed to help users scale and accelerate the impact of AI with trusted data across businesses.
The Watsonx platform is only expected to be developed further as the company moves into 2024, with Watsonx.governance having only been released in December 2023 to help businesses automate responsible, transparent machine learning and generative AI workflows.
5. Llama 2 Open-Source AI model
Launched byFaceBook (Meta), Llama 2 is an Open-Source AI model designed to further promote responsible and safe use of AI and Large Language Models (LLMs) within the industry. Trained on 40% more data than Llama 1, the previous model, Llama 2 outperforms other open source language models on many external benchmarks, including reasoning, coding, proficiency & knowledge tests.
According to Meta, Llama 2 pre-trained models are trained on two trillion tokens, doubling the context length than Llama 1. Its fine-tuned models have also been trained on over one million human annotations.
4. Duet AI AI assistant
An AI Assistant in Google Cloud, Duet AI is a collaborator available across the space to help get more done in a faster, more efficient way. Through a natural language chat interface, users can quickly chat with Duet AI to get answers to cloud questions, or receive guidance on best practices.
Duet AI is specifically trained on Google Cloud content like docs, sample code and best practices with the goal of better serving its users. It can act as a coach, thought partner and productivity booster – whilst ensuring that users and enterprises have control over their data.
3. Adobe Firefly 2 Product and an Embedded model
Created on Adobe’s Sensei platform, Adobe Firefly has been designed to assist users in expanding upon their natural creativity. Serving as both a product and an embedded model within Adobe applications, Firefly provides generative AI tools tailored to creative requirements, applications, and workflows.
Built on the technology that Adobe has produced over the last 40 years in the creative industry, Firefly is designed to make it easier for people to bring their ideas to life. The Adobe Firefly Image 2 Model is designed to be a new and improved version of the Firefly Imaging Model.
2. DALL-E 3 Text-to-Image model
DALL-E 3 represents the company’s latest version of a Text-to-Image model, offering enhanced image detail and greater accuracy in rendering faces, text, and even human hands.
When provided with an idea, ChatGPT will automatically generate tailored, detailed prompts for DALL-E 3 and work to bring the user’s idea to life. If a particular image is good, but it’s not quite right, through another prompt, users can ask ChatGPT to make tweaks with just a few words.
Just like in its previous versions, OpenAI has taken steps towards improving safety performance in risk areas such as the generation of public figures and harmful biases.
1. Gemini AI Multi-Modal reasoning
Unveiled just this week (Dec 23), Gemini is the most recently launched AI model by Google. It is currently anticipated to be one of the largest and most capable AI systems to-date. Gemini 1.0’s Multi-Modal Reasoning capabilities can already help make sense of complex written and visual information, making it skilled at uncovering knowledge in data.
According to Google, Gemini has sophisticated multi-modal capabilities, being able to master human-style conversations, language and content, as well as understand and interpret images, code, drive data and analytics and be used by developers to create new AI apps and Apps Prgm’g I/F (API s).
Please also check out our upcoming event – Sustainability LIVE Net Zero on 6 & 7 Mar 24.
BizClik is a global provider of B2B digital media platforms that cover Executive Communities for CEOs, CFOs, CMOs, Sustainability leaders, Procurement & Supply Chain leaders, Technology & AI leaders, Cyber leaders, FinTech & InsurTech leaders as well as covering industries such as Manufacturing, Mining, Energy, EV, Construction, Healthcare and Food.
BizClik – based in London, Dubai, and New York – offers services such as content creation, advertising & sponsorship solutions, Webinars & events.
Comments: any other notable Innovations 2023?
from AI Magazine 12/23 edited by Peter/CXO Wiz4.biz
For similar Info. click on Tech Trends.