Update 2: This post has been updated on March 14th, 2023 with additional information reflecting the release of the GPT-4 by OpenAI and PaLM API & MakerSuite by Google.
Update 1: This post has been updated on March 1st, 2023 with additional information reflecting the release of the ChatGPT & Whisper APIs by OpenAI.
While artificial intelligence (AI) has been around for quite some time, the recent advancements in generative AI, particularly the public research release of ChatGPT has revolutionized the way businesses operate within the span of a few months.
Now, OpenAI has released GPT-4, their ChatGPT API, meaning companies can start building full-scale applications using the model that powers ChatGPT - at 10x lower costs than previous models.
Before we dive into the details of why this is such a big deal, let's take a look at what Generative AI is and what effect it will have on businesses of all sizes around the world - and specifically, enterprises.
Generative AI is a subset of AI that is capable of creating new data or content rather than simply processing pre-existing data. The ability to generate new data or content opens up a world of possibilities for businesses. It's no wonder that generative AI is quickly becoming a hot topic in the enterprise world, crossing the chasm from being a nice-to-have for some niche startups to a tool that is enterprise-ready and that is a must-have as part of any enterprise’s technology stack to stay competitive.
The relatively nascent technology has the ability to create new revenue streams for businesses. Businesses worldwide are using Generative AI to create new content in the form of text, images, videos, and audio to be sold as a product or service to other businesses or consumers, such as in the creation of personalized content, ads, messaging, chatbots, and more. Others are using it to build additional features into their products - or building totally new products - to diversify and create new revenue streams. In both cases, the potential is excellent.
It also has the potential to massively increase speed and efficiency as well as decrease operating costs, especially for enterprise-scale companies. We will dig into these examples and more below.
One of the most significant advantages of generative AI is its ability to automate repetitive, mundane, and, let’s be honest, monotonous tasks. This frees up human resources to focus on more critical and strategic tasks such as data analysis, qualitative research, decision-making, and strategy creation.
Not only can this lead to significant improvements in efficiency, productivity, and cost savings, but also the well-being of employees.
We'd like to highlight Google's release of its PaLM API & MakerSuite because it was somewhat (putting it mildly) eclipsed by the release of GPT-4.
The PaLM API provides access to Google's large language models for various applications, including content generation, chat, summarization, and classification.
MakerSuite is a tool that simplifies the development process by streamlining prompt iteration, synthetic data generation, and custom model tuning, and it supports exporting code in commonly used languages and frameworks.
These tools aim to help developers create customized AI models while maintaining responsible and safe usage practices - something that Google has stressed is a big reason behind it's delay in releasing customer-facing applications of its LLMs.
The PaLM API also enables the generation of advanced embeddings for different applications. For scalability, Google's infrastructure supports both the PaLM API and MakerSuite, offering additional enterprise-grade support through Google Cloud Vertex AI when required.
As generative AI continues to evolve, these tools will be refined and expanded based on user feedback to better serve the development community. It remains to be seen how Google's models perform vs. OpenAI's models in production, but the initial look is promising.
Be My Eyes, a Danish startup assisting blind or low-vision individuals, is utilizing GPT-4's new visual input capabilities to develop a Virtual Volunteer™ within its app. This feature generates context and understanding similar to human volunteers, enhancing global accessibility and providing users with increased independence.
GPT-4's advanced image-to-text object recognition technology recognizes and analyzes visual information, such as the contents of a fridge, and can provide users with relevant recipes or suggestions. The technology allows for interactive conversations, delivering usable and helpful information nearly instantly.
Be My Eyes' beta-testing of the GPT-4-backed assistant has been successful, and the feature is expected to be available to users within weeks. GPT-4's conversational abilities and analytical prowess set it apart from other language and machine learning models, enabling users to navigate both the physical world and digital spaces, such as webpages, more efficiently.
Morgan Stanley, a leader in wealth management, is leveraging OpenAI's GPT-4 to revolutionize how its personnel access and utilize the company's vast knowledge base. The firm's content library contains hundreds of thousands of pages of investment strategies, market research, commentary, and analyst insights, making it challenging for advisors to locate specific information efficiently.
To address this issue, Morgan Stanley has developed an internal-facing chatbot powered by GPT-4 that searches the wealth management content repository, unlocking the company's collective knowledge. This transformational capability allows advisors to access the expertise of top strategists and analysts instantly.
The three key components of this initiative include GPT-4's ability to process and synthesize content, Morgan Stanley's extensive intellectual capital, and the expertise of the firm's financial advisors. With over 200 employees already using the system daily, the chatbot enhances the relationship between advisors and clients by enabling faster and more effective assistance. Morgan Stanley is also evaluating additional OpenAI technology to augment advisor notes and streamline client communication.
Stripe, a leading payment platform, is leveraging GPT-4 to streamline user experience and combat fraud. The company tasked 100 employees to brainstorm features and functionality for the platform using OpenAI's GPT-4. Engineers from support, onboarding, risk, and documentation teams explored how Stripe could utilize artificial intelligence to enhance features or workflows.
Stripe identified 50 potential applications to test GPT-4 and narrowed down the list to 15 strong candidates, which include support customization, answering support questions, and fraud detection.
The Stripe team is considering other potential applications of GPT, such as deploying it as a business coach to understand revenue models and advise on strategies. As GPT continues to evolve, its potential applications will expand, providing new opportunities for Stripe to enhance its platform.
Khan Academy, a non-profit dedicated to providing free, world-class education, is using GPT-4 to power Khanmigo, an AI-powered assistant that serves as a virtual tutor for students and a classroom assistant for teachers. The organization believes that GPT-4 can help address the challenges of diverse learning needs and levels among students, exacerbated by the COVID-19 pandemic.
Khanmigo is being launched as a pilot program, initially available to a limited number of participants. The AI assistant is designed to understand freeform questions and prompts, allowing for individualized and in-depth interactions with students to encourage deeper learning.
Khan Academy is also exploring GPT-4's potential to help teachers by writing classroom prompts or creating instructional materials. The organization envisions that teachers could use GPT-4 to quickly and easily tailor learning for every student and monitor their progress on Khan Academy.
GPT-4's ability to provide contextually relevant answers and explanations may help engage students by addressing their specific interests and motivations. As the technology continues to evolve, Khan Academy sees the potential to accelerate their roadmap, integrating more tutor-like abilities into their platform and offering features that were once only dreamed of.
Duolingo, a popular language learning platform, has integrated GPT-4 to introduce two new features in a subscription tier called Duolingo Max. The first feature, Role Play, offers AI-powered conversation practice, while the second, Explain my Answer, provides contextual feedback on mistakes. These features aim to improve learners' language proficiency by facilitating implicit learning, or learning through repeated use of vocabulary and grammar in various contexts.
Initially available for Spanish and French courses, Duolingo Max has significantly impacted the company's engineering process by allowing rapid prototype development and enabling the team to focus on testing and honing data sets. Duolingo plans to expand these features to more languages and introduce additional innovations to maintain its leadership in language learning.
Iceland has partnered with OpenAI to preserve the Icelandic language by using GPT-4. The government aims to improve GPT-4's abilities for Icelandic and create resources for preserving other low-resource languages. A team of 40 volunteers is training GPT-4 on proper Icelandic grammar and cultural knowledge using Reinforcement Learning from Human Feedback (RLHF), making it more feasible for low-resource languages to replicate the process.
The ultimate goal is to enable GPT-4 to power complex and creative applications in Icelandic, allowing Icelandic companies to deploy GPT-4 in their interactive applications. Miðeind's voice assistant app, Embla, will use GPT-4 as its backend for conversations in fluent Icelandic and translations to other languages. This partnership also supports Icelandic companies in using Icelandic chatbots on their websites instead of relying on English-speaking ones.
Salesforce unveiled Einstein GPT, the first generative AI for CRM, to transform customer experiences across various sectors. The technology integrates with OpenAI for out-of-the-box AI capabilities and real-time adaptive content generation. Salesforce Ventures is launching a $250 million Generative AI Fund to invest in promising startups and promote responsible AI. The ChatGPT app for Slack, built by OpenAI and Salesforce, offers AI-powered conversation summaries, research tools, and writing assistance.
Instacart's "Ask Instacart" feature, set to launch later this year, will use ChatGPT alongside Instacart's AI and product data to offer inspirational, shoppable answers to customer inquiries about food.
With access to more than 75,000 retail partner stores, customers will be able to receive personalized recommendations for open-ended shopping goals, such as finding healthy lunch options for their kids or making great fish tacos.
Snapchat's new feature, My AI for Snapchat+, offers a customizable chatbot experience for its 750 million monthly users.
Powered by the ChatGPT API, My AI not only provides friendly recommendations but also boasts the ability to write a haiku for friends in mere seconds.
Shopify's Shop app, used by 100 million shoppers, has introduced an AI-powered shopping assistant that offers personalized recommendations based on user requests.
Powered by ChatGPT API, the shopping assistant can quickly scan millions of products to help shoppers find what they're looking for or discover new products.
Quizlet, the global learning platform with 60 million students, has partnered with OpenAI for three years to enhance vocabulary learning and practice tests using GPT-3. This moves comes as a clear challenge to the traditional educational system that has by and large rejected the use of ChatGPT or any AI tools.
Now, with ChatGPT API, Quizlet is introducing Q-Chat, an AI tutor that adapts to students' learning progress and provides relevant study materials delivered through an engaging chat experience.
Speak, the fastest-growing English app in South Korea, is using the Whisper API to power an AI speaking companion product for language learners worldwide.
Whisper's human-level accuracy provides open-ended conversational practice and highly accurate feedback, helping learners build their path to spoken fluency.
This is the clearest example of how reliable, enterprise-ready, large-scale-ready, and secure OpenAI’s various technologies are at this point in time.
If you ever had a doubt whether they will fulfill your needs, this should alleviate any and all of these doubts.
With Microsoft's Azure OpenAI Service now Generally Available (GA), companies may request access to some of the most cutting-edge AI models in the world. With enterprise-grade security and dependability, Azure OpenAI Service made its debut in November 2021 with the goal of making large-scale generative AI models more accessible, especially to enterprises.
Since then, a variety of use cases, including customer service and extracting insights from data through search, data extraction, and categorization, have made use of it.
The collection of OpenAI API models' fundamental computing power, Microsoft Azure, offers companies and developers high-performance AI models at a production scale.
Azure is the only global public cloud that offers AI supercomputers with large-scale capabilities and is used by the leading AI companies in the world. Azure ranks in the Top 15 of the Top 500 supercomputers worldwide and is the highest-ranked global cloud services provider.
Building enterprise-scale applications on Azure has never been easier. As a Microsoft Gold Partner for the past 13 years, we are proud to be helping our clients design and develop cutting-edge applications on Azure. Get in touch to learn more about our Azure consulting services.
In a surprise press conference hosted on February 7th, 2023, at 10AM Pacific Time, Microsoft CEO Satya Nadella announced OpenAI technology’s integration with Bing Search and Edge.
The new Bing and Edge integrate search, browsing, and chat into one seamless experience, delivering better search results, complete answers, and a new chat experience. The new Bing uses a next-generation OpenAI model, and Microsoft's proprietary Prometheus model to provide more accurate, relevant, and targeted results.
The AI model has also been applied to the core search ranking engine, leading to a significant improvement in search relevance. The new Edge browser has been updated with AI capabilities and a new look, and now includes chat and compose functionalities. The unified experience of search, browser, and chat is designed to provide a completely new way to interact with the web.
The pandemic has led to a 252% increase in weekly time spent in meetings, making it essential to work smarter. Microsoft Teams is incorporating new AI-powered capabilities across its consumer and enterprise products, including Microsoft Teams Premium.
Teams is infusing AI throughout the meeting experience to help users be more productive in new ways. With intelligent recap, users will receive automatically generated meeting notes, recommended tasks, and personalized highlights, even if they missed the meeting.
With Viva Sales, Microsoft is launching a brand-new generative AI-powered experience. By automatically recommending customizable material, this new experience will let merchants engage with prospects and consumers more successfully.
They will be able to concentrate on important tasks like forming deep connections, establishing trust, and developing long-lasting relationships since they will have more time, thanks to this. The seller may choose the option that best suits their needs, and Viva Sales will produce suggested email text for a number of circumstances.
The seller can then review and change the generated reply to their satisfaction. The Microsoft Graph, which gives users access to people-centric data and insights in the Microsoft Cloud, as well as Conversation intelligence, will be coupled with this new GPT function.
Intercom has introduced its AI-powered features for customer service.
Des Traynor, Co-founder and Chief Strategy Officer at Intercom, believes this technology will be a major change for the customer service industry and is excited about exploring what's next.
Emplifi, a leading customer experience platform serving over 8,400 brands - including Delta, Ford, and McDonald's, has launched Emplifi AI Composer, an AI solution that generates ready-to-publish social media content for businesses.
The tool is integrated into Emplifi's Social Marketing Cloud and utilizes OpenAI's GPT-3 framework. The solution helps businesses increase operational efficiency and productivity by automating social media content generation.
Emplifi’s AI Composer prompts social media teams to input their message instructions and select customizations, then modifies and generates the content using the GPT-3 framework. Features include omnichannel flexibility, human-led content, and real-time performance analytics.
NICE has integrated its cloud-based knowledge management solution, CXone Expert, with OpenAI's ChatGPT technology to improve self-service customer support.
The integration creates personalized conversational AI experiences for customers and provides them with immediate, accurate, and human-friendly responses. By combining CXone Expert's content retrieval and conversational search with ChatGPT's natural language processing, companies can deliver improved self-service experiences that are more human-like, reducing the need for customer transfers or call-backs.
We are working on compiling all the companies and the products they're bulding using Anthropic and their AI Chatbot Claude. In the meantime, you can see how the companies are using Claude, Anthropic's ChatGPT challenger, to help companies like Notion, Quora, DuckDuckGo, and many more!
It’s hard to get the needle moving in large organizations with many moving parts. Hiring a full in-house team is not ideal, especially if you’re still exploring who you need on that time, and it takes a long time to get going. To keep up with your competition, consider working with a technology consultancy and tech partner to . Reach out to us directly, and our team can help walk you through how to build apps on Azure using the Azure OpenAI Service.
While these applications and the possibilities we will highlight below are impressive, implementing these use cases often need a conscious, intentional decision and lots of additional effort to use these solutions effectively.
Since these Large Language Models (LLMs) are trained on data that is not “live” or up to date, they need to be fine-tuned using additional training data or in this case, by using a specific type of training called Embeddings. This allows the models to access real-world data and base answers on that data.
If your department or company is looking into creating custom software built on top of LLMs, consider reaching out to an experienced technology partner to help you get started.
Generative AI has the potential to disrupt the way data analysis and decision-making are carried out.
By analyzing patterns in historical data and generating insights, generative AI models can identify trends and potential opportunities for growth.
These models can also automate decision-making by predicting the potential outcomes of various strategic moves, allowing software engineering leaders to make data-driven choices confidently, saving time and effort in the process.
In the realm of customer service and support, generative AI can be employed to improve the entire experience for both customers and agents. Using natural language understanding and generation technologies, AI can automate the handling of routine customer queries, significantly reducing response times while maintaining a high level of accuracy and personalization.
Generative AI can also be used to answer questions better using existing support content, such as FAQs and troubleshooting guides. The use of Large Language Models (LLMs), which are capable of understanding intent really well, is central to this process.
They can answer ad hoc questions with answers that bring information from multiple sources, making it easier for customers to get a better support experience by helping them find the answers they need more quickly.
Additionally, AI can assist human agents by suggesting the best approach to handle complex issues, taking into account the customer's needs, preferences, and available solutions.
Generative AI has the potential to revolutionize supply chain management and logistics by optimizing resource allocation, route planning, demand forecasting, and delivery schedules, which can help businesses save time and money.
AI can identify the most efficient paths for transporting goods based on factors like traffic, weather conditions, and fuel consumption. Furthermore, it can make recommendations for adjusting inventory levels, thus helping to prevent stock-outs and overstock scenarios, which can help businesses stay competitive and respond more effectively to changes in the market.
Generative AI has the potential to make financial services and accounting more efficient and precise. AI can automatically analyze enormous amounts of financial data, detecting anomalies and fraudulent activities with higher accuracy than manual reviews. It can also be used to generate new financial models, which can help businesses make more informed investment decisions.
Additionally, generative AI can be used to analyze financial data and identify patterns & trends, and can automatically generate financial reports & forecasts, which can help businesses make better decisions and stay ahead of the curve.
Generative AI has the potential to revolutionize product development and design. Generative AI can be used to generate new product designs, which can save time and money for businesses that rely on product design as a core part of their operations.
In the field of product development and design, generative AI can help analyze consumer data and generate new product ideas based on preferences, trends, and functionality, which can help businesses stay ahead of the curve and meet the evolving needs of their customers. By iterating quickly through multiple design variations, generative AI models can significantly reduce the time and cost involved in the design process while enhancing overall product quality.
When it comes to content creation, businesses can use generative AI to generate new marketing or advertising content, which can help businesses reach new customers and grow their customer base.
Generative AI can be applied within human resource management and recruitment to automate time-consuming tasks and enhance decision-making.
AI-driven systems can assess candidates' resumes, qualifications, and skills to generate shortlists and make recommendations for interviewing based on the job requirements. Generative AI can also be used to generate new recruitment strategies, which can help businesses attract and retain the best talent.
Moreover, generative AI can be used in performance management and training, identifying skill gaps and suggesting personalized training programs for employees.
Generative AI can boost the effectiveness of marketing and advertising campaigns by personalizing content, messages, and targeting recommendations. By analyzing customer behavior, preferences, and trends, generative AI can create tailored marketing materials that resonate with target audiences.
For example, generative AI can be used to generate new marketing content and campaigns, which can help businesses reach new customers and grow their customer base.
Generative AI can also analyze consumer data and identify patterns and trends, which can help businesses make better marketing and financial decisions. Finally, AI can automate the process of A/B testing for different ad formats, channels, and creative designs, helping marketers maximize ROI on their campaigns.
Generative AI has the potential to revolutionize manufacturing and production operations by making them more efficient and adaptable. AI can optimize production schedules, manage resources, and reduce waste, all while maintaining high-quality outputs.
Generative AI can also be employed in predictive maintenance, utilizing sensor data to identify potential problems in machinery before they become costly failures.
Additionally, generative AI can be used to generate new product designs, which can help businesses stay ahead of the curve and meet the evolving needs of their customers.
The business sector is not the only domain that generative AI has revolutionized. Generative AI also has immense potential in healthcare and medicine, including applications in drug discovery, patient care, and medical diagnostics.
For example, AI can analyze chemical structures, patient data, and clinical trial results to generate potential drug candidates. Additionally, generative AI models can empower diagnostic tools, enabling more accurate and faster detection of diseases, ultimately improving patient outcomes and overall healthcare efficiency.
Generative AI can also be used to improve healthcare and medicine. For example, generative AI can be used to analyze patient data and identify patterns and trends, which can help doctors make more informed treatment decisions.
While generative AI has many advantages, there are also ethical considerations that need to be taken into account. For example, there are concerns about the impact of generative AI on jobs and the workforce, as well as concerns about the potential misuse of generative AI. Another concern is the presentation of false information as factual.
There are also questions about the responsibility of businesses and organizations when it comes to the use and implementation of generative AI. Questions are being raised about whether specific jobs that are currently an essential part of the workforce will become obsolete, such as content creation and customer service jobs. It is important for businesses to consider these ethical considerations and take steps to mitigate any potential negative impacts.
Generative AI also risks reflecting existing human and social biases. If the data collected by AI is the data that is at a statistical majority, then the data is also biased and will solely reflect that majority.
Research also shows that AI is going beyond bias and is sometimes fabricating information and presenting it as the truth (sometimes referred to as hallucinating). If left unchecked, this could lead to dangerous results when used in the fields of medicine, law, or engineering, to name a few.
As the use of generative AI becomes more widespread in the enterprise, it is important to consider the potential impact on privacy and security. Generative AI models are trained on large amounts of data, and they can generate new data that is similar to the training data. This raises concerns about the potential for misuse or abuse of this data, such as the generation of deepfake videos or the creation of fake news.
In order to mitigate these risks, businesses need to be transparent about their use of generative AI and ensure that they have robust privacy and security policies in place. This may include measures such as data encryption, access controls, and regular security audits. It is also important for businesses to be aware of the legal and regulatory requirements related to the use of generative AI and to ensure that they are in compliance with these requirements.
The future of generative AI in the enterprise is looking bright, and more and more businesses will begin to adopt this technology in the coming years. The field of generative AI is constantly evolving, and new applications and use cases are being discovered all the time.
In the future, we can expect to see even more innovative ways in which generative AI is being used in the enterprise. For example, generative AI can be used to create realistic virtual environments for training and simulation, generate new financial models and forecasts, and improve logistics and supply chain management.
However, it is important for businesses to consider the potential impact of generative AI on the workforce and take steps to mitigate any negative impacts. This may include retraining workers for new roles or investing in education and training programs to help workers adapt to the changing landscape.
While the benefits of generative AI are clear, there are also challenges that need to be overcome in order to implement this technology in the enterprise successfully. Some of the biggest challenges include the need for specialized expertise and the high cost of implementation.
Due to the conversational nature of the prompts that people use while communicating with the AI models and the fact that this data is used to train future models further, there are also concerns about data privacy and security. However, with the right approach, businesses can overcome these challenges and reap the benefits of generative AI.
Generative AI is one of several different types of AI techniques that are used in the enterprise. Other types of AI include supervised learning, unsupervised learning, and reinforcement learning. Each of these techniques has its own strengths and weaknesses, and they are used for different types of tasks.
Supervised learning is a type of AI that is used to train models to classify or predict outcomes based on input data. This is the most common type of AI used in the enterprise, and it is used for a wide range of tasks such as image recognition, natural language processing, and predictive analytics.
Unsupervised learning is a type of AI that is used to identify patterns and relationships in data without the need for labeled training data. This is typically used for tasks such as anomaly detection and cluster analysis.
Reinforcement learning is a type of AI that is used to train agents to make decisions based on feedback from the environment. This is typically used for tasks such as game playing and robotics.
Generative AI is different from these other techniques in that it is used to generate new data rather than just classify or predict outcomes. This makes it a powerful tool for tasks such as image generation, language generation, and video generation.
Generative AI is not an isolated technology, but it can be integrated with other technologies to achieve new capabilities.
For example, generative AI can be integrated with computer vision and natural language processing to create new forms of multimedia content. It can also be integrated with simulation and modeling to create new virtual environments for training and testing.
Generative AI can also be integrated with blockchain technology to create new forms of decentralized applications.
Conclusion
In conclusion, Generative AI is a powerful tool that has the ability to create new forms of data, products, and services. It has the potential to transform a wide range of industries and create new revenue streams for businesses. As the use of generative AI becomes more widespread in the enterprise, it is important for businesses to be aware of the potential impact on privacy and security and to ensure that they have robust policies and procedures in place.
.svg){kind=small}
