Despite the worldwide buzz around generative AI, this technology isn’t brand-new. Generative artificial intelligence was introduced in the 1960s in chatbots and, until 2014, remained in the background. 2014 has become a tipping moment for this technology due to the introduction of generative adversarial networks (GANs) - a machine learning algorithm with the help of which generative AI can create authentic text, images, and audio and video of real people.
This newfound capability has opened up many opportunities, including rich educational content, design and creativity, simulation and modeling, personalization, and automation. Nowadays, generative AI is rapidly becoming a game-changer for almost every industry. With its help, businesses can generate creative and personalized content for their customers, researchers can simulate and model complex systems, and artists can create novel and innovative works.
Source: https://www.enterpriseappstoday.com/
The technology behind generative AI is advancing at an unprecedented pace, making it an exciting and rapidly developing field with immense potential. This article will explore the basics of generative AI, how it works, and its potential applications across the construction industry.
Generative AI Market Overview
Generative AI has become a breakthrough for almost every industry. This technology uses complex algorithms based on machine learning models to produce text and images, program code, artwork, poetry, and numerous other tasks. Generative AI creates a targeted summary after searching the necessary information through various legal sources, generating code and text in real-time and offering a wide range of functions, snippets, and fully working modules. This technology is expected to save countless hours of human work and research and be a game-changer for numerous industries in the coming years.
Advancements in technology, task automation, and increasing demand for personalized content are the main driving factors fuelling the generative AI market growth. Along with the development of virtual environments like the metaverse and applications across various industries, including e-commerce, gaming, and healthcare, these factors play a crucial role in generative AI market development and the evolution of AI and deep learning technologies in general.
According to recent statistics, the global generative AI market is expected to reach $110.8 billion by 2023, growing at a CAGR of 34.3% during the forecast period. The COVID-19 outbreak has also positively impacted the generative AI market, as many organizations adopted Machine Learning and Artificial Intelligence in response to the pandemic.
The global generative artificial intelligence market is segmented into:
- Components;
- Regions;
- Technologies;
- End-users.
Depending on the component, it is divided into software and services, with the software segment dominating the generative AI market in 2022. This trend is expected to continue due to the surge of NLP advancements promoting market growth. However, the services segment is projected to witness the most significant growth over the forecast period due to the increased usage of cloud-based generative AI services that provide scalability, flexibility, and cost-effectiveness for users.
As for regional distribution, the AI market dominated North America in 2021 and is projected to keep its position over the next few years. However, Asia Pacific is expected to witness noticeable growth during the forecast period due to rapid business digitization and increased load on cloud networks and data centers. The introduction of artificial intelligence helps organizations enable civil society members to be responsible and informed users of AI devices.
Currently, the global generative AI market is dominated by key players such as:
- Adobe, Inc.;
- Amazon Web Services, Inc.;
- D-ID, Genie AI Ltd.;
- Google LLC;
- IBM Corporation;
- Microsoft Corporation;
- MOSTLY AI Inc.;
- Rephrase.ai, Synthesia.
Many of these generative AI market giants offer solutions to various applications, such as super-resolution, text-to-image, image-to-image, and others. These players are also researching generative AI technology for advanced video and image resolution and face aging. For example, the American company Tesla is developing autonomous algorithms using data from car sensors. Also, IBM Watson's Tone Analyzer uses natural language processing to analyze the tone of written text. Companies use this generative AI tool to improve customer engagement by identifying and addressing negative or potentially problematic language in customer interactions.
For now, less than 1% of online content is generated by artificial intelligence. Within the next ten years, this percentage is predicted to grow to 50%, resulting in half of all online content being fully generated or augmented by AI. Generative AI has broad applications across numerous areas, from media and communications to life sciences and beyond, offering, in most cases, lower costs and higher value.
Generative AI, OpenAI, and ChatGPT. What’s the difference?
As ChatGPT emerged onto the scene in November of 2022, it quickly became a buzzword, sparking a wave of excitement among the public and prompting Microsoft and Alphabet to hurry up and release their own products, as they believed this technology had the potential to revolutionize the way people work. The chatbot attracted over a million users in the first five days and hit a mark of 100 million monthly active users in just two months, becoming the fastest-growing consumer app ever. People worldwide were amazed by its ability to write essays, code, and even jokes, using language that was so natural it was hard to believe that artificial intelligence generated it.
Source: https://subredditstats.com/
However, ChatGPT is not the only innovative technology that keeps capturing people's attention nowadays. Alongside ChatGPT, other tools such as Midjourney and DALL-E had also begun producing stunning “AI art” works while AIVA was easily composing music. With the advent of ChatGPT, a new type of content creation has come to the fore - generative AI. Let’s find out the difference between generative AI, OpenAI, and ChatGPT and check what pros and cons these technologies can give us.
Generative Artificial Intelligence
Like other artificial intelligence forms, generative AI learns to take actions based on past data. Unlike other forms of AI that categorize or identify the data, this technology creates completely new content - text, image, or computer code based on that training. Among the main characteristics of generative AI are:
- Unsupervised learning. In contrast to traditional AI models requiring large amounts of labeled data to train, generative AI can learn from raw data and discover patterns on its own.
- Creativity. Generative AI can create new and unique output, which can be used to generate text, music, or art.
- Adaptability. This form of AI can adapt to new data and situations, which allows it to create new content tailored to specific contexts or needs.
- High-dimensional data. Generative AI creates high-dimensional data, such as videos or images, that require numerous parameters to represent the data accurately.
- Diversity. Generative AI can generate various outputs by sampling from the stochastic probability distribution obtained during training, which allows it to create a wide range of results for a given input.
- Probabilistic modeling. This AI form uses probabilistic modeling techniques to create new content. It includes modeling the probability of different outcomes based on patterns and relationships learned from the training data.
- Computational intensity. Generative AI models require high-performance computing resources like GPUs to produce high-quality outputs.
Source: https://www.idc.com/
There are three main generative AI models:
- Generative Adversarial Networks (GANs). GANs consist of two neural networks - a generator and a discriminator that work together to generate new data. The generator creates new samples while the discriminator evaluates whether they are real or fake. Over time, the generator learns to create increasingly realistic samples.
- Variational Autoencoders (VAEs). This generative AI uses neural networks to learn the underlying distribution of a given dataset. VAEs can generate new data samples that are similar to the original dataset.
- Transformer-based models. Transformers represent neural network architecture that uses self-attention mechanisms to generate new data. They are highly effective at natural language processing tasks, such as text generation, and have been used to create realistic and coherent text samples, images, and music.
OpenAI
OpenAI is an artificial intelligence research laboratory that consists of scientists and researchers who are dedicated to developing and promoting AI that is safe, beneficial, and accessible to everyone. This organization was started by Elon Musk in 2015, but he stepped down from his position in 2018 due to potential conflicts with Tesla. Currently, OpenAI is headed by its President and Chairman, Greg Brockman, Chief Scientist Ilya Sutskever, and Sam Altman, its CEO, with Elon Musk remaining one of the main donors of the company.
OpenAI creates and promotes various forms of AI technologies, including generative AI, but it is not a specific type of AI technology. The main goal of OpenAi is to create AI that would benefit all and ensure that it is developed and used safely, transparently, and ethically.
ChatGPT
ChatGPT (Generative Pre-trained Transformer) is a state-of-the-art language model developed by OpenAI that uses deep learning techniques to generate human-like responses to various queries. Its main characteristics include the ability to understand natural language and generate coherent and contextually relevant responses in a conversational setting. ChatGPT is an example of generative AI, but it is not the only type of generative AI technology. Among other types of this technology are:
- Image Generation. AI models like GANs (Generative Adversarial Networks) and Variational Autoencoders.
- Music Generation. AIVAs (Artificial Intelligence Virtual Assistants) and Magenta.
- Video Generation. StyleGANs (Style-based Generative Adversarial Networks) and VQ-VAE (Vector Quantized Variational Autoencoder).
- 3D Object Generation. Deep Dream and GANs.
Source: https://www.cnbc.com/
ChatGPT has many potential applications, including customer service, content creation, personal assistants, language translations, and mental health support. These are several examples of the many potential use cases for ChatGPT, and as the technology continues to evolve, new use cases are likely to emerge.
Pros, Cons, and Application of Generative AI
Generative AI has great potential to transform many industries by generating diverse and original content at scale, automating creative tasks, and improving efficiency and productivity. However, it also comes with some challenges, such as ethical concerns around bias and privacy, the need for significant computing resources, and the possibility of job loss in specific industries. In the table below, we’ve collected the advantages and disadvantages of generative AI and its most promising applications in various fields.
| Pros | Cons | Applications | Examples |
|---|---|---|---|
| Generates original and diverse content quickly and at scale | Can generate biased or inappropriate content | Content (text, image, music, video) creation | ChatGPT |
| Automates creative tasks and reduces workloads for humans | May lead to job loss in some industries | Personalized assistants and chatbots | Google Duplex for conversational AI |
| Enhances efficiency and productivity. | Requires significant computing resources | Customer service and support | Autodesk Dreamcatcher for design optimization |
| Makes processes fast and inexpensive | May raise ethical concerns around privacy, bias, and safety | Healthcare (e.g., mental health support) | Woebot for mental health support |
| Learns from user data and preferences to create personalized content | Lack of human control can lead to errors, mistakes and unforeseen consequences | Language translation. | DeepL for language translation |
| Can simulate complex systems, such as weather patterns or financial markets | May create content that violates the intellectual property rights of others | Scientific research and data analysis | Eureqa for scientific modeling |
| Generates synthetic data that can be used to augment existing datasets and enhance the performance of machine-learning models | Can undermine public trust in the authenticity of content | Education and training | Artie for interactive educational content creation |
| Allows companies to explore new ideas and concepts | Can be used for malicious purposes | Design and fashion | Prisma for artistic image generation |
| Creates content in various formats | Can generate fake content that can be difficult to distinguish from genuine content | Gaming and entertainment | DALL-E for image generation from textual input |
Conversational AI (ChatGPT) in the Construction Industry
The global construction market reached the mark of $12.74 trillion in 2022 and is estimated to attain a value of $18.59 trillion by 2028, growing at a CAGR of 6.5%. Different technologies like DataCAD, REVIT, ArchiCAD, NavisWorks, CADKEY, and StruCad are currently being actively used to save time and costs during the entire construction process, pushing the market growth.
The global AI in the construction market was valued at $398.6 in 2018 and continues to evolve. Thanks to advances in artificial intelligence, construction companies can now predict and prevent hazards at job sites with greater accuracy and safety measures. It has boosted the adoption of AI technology, especially for predictive maintenance and risk management solutions. Moreover, AI plays a crucial role in developing Industrial Internet of Things (IIoT) applications that help project managers achieve higher levels of efficiency and reduce labor costs. Growing demand for IIoT solutions drives construction companies to invest in AI-based technologies, leading to a move towards full automation and more innovative job sites.
Source: https://www.mordorintelligence.com/
In this paragraph, we’ll consider the benefits of Conversational AI (ChatGPT) in the construction industry and check its potential to cope with the main construction issues. We’ve also collected some solutions that can be built with generative AI to improve efficiency and accuracy in the construction industry.
Conversational AI overview
Chatbots, also known as conversational AI, are software agents that can mimic human conversation and provide automated assistance to users. They have become popular due to the increasing demand for efficient and convenient customer service in various industries such as e-commerce, finance, and healthcare. According to report Conversational artificial intelligence in the AEC (architecture engineering and construction) industry: A review of present status, challenges and opportunities
Popularity in the development of Conversational AI systems is a result of advancements in machine learning, the development of graphics processing units (GPU), and the involvement of big companies. The success of these Conversational AI systems can be related to the improvement of human–computer interactions they brought about. It is not surprising that the Conversational AI global market size is expected to grow at a compound annual growth rate of 21.9 % moving from 4.8 billion USD in 2020 to 13.9 billion USD in 2025.
There are two types of chatbots: domain-specific or goal-oriented chatbots, which focus on a specific task such as ordering food, booking a flight, or scheduling an appointment, and non-task-oriented or open-ended chat systems, which are for chit-chat and do not have a specific objective.
Both types of chatbots are designed to process and understand human language input, but there are some key differences between them. Non-task-oriented chatbots, also known as virtual assistants, use AI and big data technologies to provide users with a general answer through question-and-answer-like conversations between humans and machines. These chatbots can cover almost all kinds of topics, and they are mostly based on a deep learning model that is pre-trained on a large corpus of text data. Examples of virtual assistants include Siri, Alexa, Cortana, Google Assistant, and Bixby. In contrast, task-oriented chatbots are designed to perform a specific task, and they do not require as massive a training dataset as non-task-oriented chatbots. Most of them use machine learning algorithms or rule-based methods to understand the context. A new generation of non-task-oriented chatbots, such as ChatGPT, can even understand complex questions and generate complete text on a wide range of topics, instead of just providing simple answers with a few sentences. These chatbots are trained on a massive dataset sourced from the internet, which includes a diverse range of text.
Conversational AI is expected to change and support business models, supply chains, logistics, customer relationships, and improve productivity across industries. Applications of Conversational AI systems are increasing in e-commerce (to attend to customers), in the health care sector (for providing information to cancer patients, sex education, professional training, and chronic diseases), in travel, tourism, and hospitality. However, the application of Conversational AI in the AEC industry is still in the infancy stage compared with other sectors.
ChatBots Benefits in the Construction Sector
ChatGPT as part of Conversational AI stands out as a revolutionary version of the popular language model GPT, making it ideal for automating various processes in the construction industry. ChatGPT can be used to automate multiple tasks like answering frequently asked questions, scheduling appointments, and providing real-time updates on project progress. It can also assist in the design process by generating detailed specifications and plans based on high-level input from architects and engineers and creating logical and accurate task breakdowns for the construction project.
The use of ChatGPT in a construction project is not limited to scheduling and cost management but can cover any other project management perspective like quality, health, safety, security, and environment (HSSE), sustainability, and communications, among others. Scientists from New York University Abu Dhabi provided the AI with a floor plan, a list of tasks, and a description of the scope of work and found that ChatGPT could help with basic project management.
Opportunities for Conversational AI in architecture engineering and construction (AEC) industry
The most promising benefits of ChatBots in the construction industry include:
- Real-time info about machinery and team performance;
- Sharing photos and progress at the construction site;
- Real-time activity tracking;
- Easy access to drawings and location information;
- Notification of urgent needs and requirements;
- An easy exchange about contractors, warehouses, and materials;
- Publishing a daily progress report;
- Notifications to site engineers;
- Real-time activity tracking;
- Timely procurement and contractor management;
- Quick update on construction planning tools;
- Easy multidisciplinary coordination;
- Easy access to construction documentation and drawings.
Due to a shortage of domestic labor and the global pandemic, there is a need for a digital solution that can provide construction site workers, especially site managers, with information efficiently to aid in their daily management tasks. ChatBots can manage workflows to increase location-based interaction with equipment and engineers for planning and cost control. Today, AI-based chatbots are among the essential assistants to engineers on construction sites, allowing them to effectively manage the relationship between design, implementation and use of the equipment and significantly increase overall productivity.
Construction Sector Challenges and Conversational AI Solutions
For now, coordination is one of the most significant challenges in the construction sector. Different parties are working on the same tasks using various communication channels, which creates unnecessary complexity and leads to conflicts of responsibility with unexpected changes in workflow that increase rework and errors and reduce productivity. The TOP 3 causes of miscommunication at construction sites are:
- Unresponsiveness between project team members.
- Inability to collaborate effectively among project stakeholders.
- Lack of an effective platform where stakeholders and project participants can easily communicate and exchange information about the project.
In addition, there are challenges in real-time data collection in the construction sector. To succeed in project management, it is essential to manage construction-related information. To date, instant messaging applications such as Slack, WhatsApp, and WeChat are used to share daily construction information, but extracting and integrating this data for daily reports is still made manually by contractors because input data is usually unstructured and applications for instant messaging does not usually work with a construction management system. In the section below, we’ve collected the main pain points of the construction industry and solutions that can be built with ChatGPT.
| Challenges | Potential Solutions with ChatGPT |
|---|---|
| Staff shortage | Chatbot for recruitment processes, including screening candidates and scheduling interviews. |
| Tasks coordination | Bot for project management processes like scheduling, quality, health, safety, security, and environment (HSSE), sustainability, and communications. |
| Change and issue management | Chat for change management, which would help to identify problems and assign responsibility for their solution. |
| Technology adoption | Chatbot for training and onboarding new technologies that provide tutorials and answer questions. |
| Real-time data collection | Bot for real-time monitoring and analysis of construction data, identifying areas for process improvement and optimization. |
| Collaboration among project stakeholders | Chatbot for facilitating communication and collaboration among project stakeholders, including sharing project updates and resolving conflicts. |
| Daily report composing | Bot for collecting and synthesizing data from various sources into comprehensive daily reports. |
| Decision making | Chatbot for analyzing construction data and making data-driven decisions, including identifying trends and making recommendations. |
These are just a few examples of potential solutions to construction problems. The use of ChatGPT in the construction sector is still a relatively new and growing field, and there may be other innovative ways that can be applied to address the pain points faced by the industry.
Wrapping Up
The future of generative AI is exciting and full of possibilities. Advancements in natural language processing will enable machines to produce more human-like language and improve the accuracy of text generation. In addition, enhanced visual content generation will lead to better image and video manipulation and creation, while real-time feedback will allow for quicker and more efficient content creation.
Generative AI will continue to be used for personalized content creation and integrated with other technologies, opening up new creative possibilities. However, ethical considerations such as transparency, accountability, and avoiding bias still need to be addressed.
The use of ChatGPT in the construction industry presents a promising opportunity for improving efficiency and accuracy. With its ability to automate various processes, such as answering frequently asked questions and scheduling appointments, this chatbot can help significantly reduce the workload for construction professionals and increase productivity. Moreover, ChatGPT can assist in the design process by generating detailed specifications and plans based on high-level input from architects and engineers.
The potential benefits of ChatGPT go beyond simple task automation, as it can also provide real-time updates on project progress, ultimately helping to ensure timely project delivery. ChatGPT has every chance to revolutionize the construction industry and make it more efficient, accurate, and cost-effective.