BLOG
BLOG
We believe in AI and every day we innovate to make it better than yesterday. We believe in helping others to benefit from the wonders of AI and also in extending a hand to guide them to step their journey to adapt with future.
Generative AI has shaken the netizens with its magic-like capabilities. Everyone has to share an experience of how they got amused by Generative AI (GAI) models like GPT3, Stable Diffusion, etc. Anyone would agree that the latest advancements in Generative AI can potentially disrupt not just a couple of industries but the whole world itself. It’s like the early stages of the inception of the internet. History has taught us multiple times that those who embrace the change will thrive, and the rest will eventually fade. In these deterministic years, the priority for anyone and any business should be to discover how to embrace GAI and leverage it for growth.
In this article, I want to focus less on the common use cases and applications of GAI models —for example, using GAI for automated blog writing, poster creation, etc. There are 100s of articles on the internet that explain the heads and tails of the simple use cases that are very obvious even for a kid. I want to focus more on high-value-driven use cases. Growth opportunities that are not obvious.
Generative AI is a category of AI that focuses on ‘generating something’ that didn’t previously exist. Examples of AI models that generate content like text, images, audio, and video are now familiar to us. ChatGPT, Dalle, and Stable Diffusion are some popular examples. Generative AI uses algorithms to generate, manipulate, or synthesize data, often visual content or human-readable text.
Its spreading like a wildfire,
OpenAI’s ChatGPT has reached an estimated 100 million monthly active users just two months after its launch. It is the fastest-growing consumer application in history. To give you a perspective, TikTok and Instagram took nine months and 2.5 years to onboard 100 million users. When considering the inherent value delivered by the product, the humungous user acquisition rate is very reasonable. In an update from OpenAI in Sep 2022, they said that more than 1.5 million users use DALL-E daily, creating over 2M images. Midjourney, a GAI model similar to DALL-E, receives about 4M monthly website visitors. Another text-to-image model, Stable Diffusion, is reportedly used by more than 10 million people. All these user acquisition figures are attained in an alarmingly short period. Clearly, the general public welcomes the technology, and undoubtedly, GAI is the FUTURE.
As the world is quickly embracing GAI, business owners and stakeholders must ask themselves a billion-dollar question. “Should I make a move now or wait?” History has taught us that those who resist change, especially those resisting embracing technological advancements, have nothing to expect but eventual death. That’s what happened to Kodak, Nokia, and numerous other examples. The best time to brainstorm on how your business can embrace AI was many years ago. The second best time is right now. Despite the technological advancements, the industrial adoption of AI technology is still in its early stages. Now, the question you may have is, how exactly can I leverage this opportunity and ride the wave with an early mover advantage?
In this section, let’s focus more on high-value-driven use cases. Growth opportunities that are not obvious. Here are some business growth opportunities of GAI;
Generative process automation is obvious. You can use GAI models to automate content generation. The obvious use cases are automating generative processes like writing a letter or a blog, summarising a business proposal, analyzing a legal document, creating posters and decorative photos, etc. GAI can complete such tasks in minutes, which could otherwise be a week-long process for your best employees. Let’s dig deep into the non-obvious opportunities.
Generative AI’s capability is not just limited to creating beautiful images or writing content copies. If you use it correctly, you have one of the best tools to accelerate your business growth. For example, suppose your company is launching a new product. How long would it take for your marketing team to develop a launch plan? Whatever it may be, GAI models can augment your employees to complete the job in an hour maximum. By using a carefully designed set of prompts in an order, you can get the GAI model to perform tasks like;
The above is just an example. You can automate a series of processes for a lot of activities. Another example would be product development. From creating user stories to creating code and code quality analysis, you can simply use a list of prompts to complete the job quickly. To get started, you can hire expert prompt designers or consultants to design the series of prompts to automate the business process you want to automate with a language model or GAI.
The above opportunity involves designing a set of ordered prompts to automate business processes. But what if you do not get the desired output for a particular prompt?
Consider the example above. Suppose you require action items in a tabular form—a table with an action item and a schedule to execute it. But, the response you got from the model is a descriptive paragraph. This situation is where you can use prompt engineering.
You can design and optimize text-based instructions for language models to generate accurate and relevant outputs via prompt engineering. Prompt engineering improves language models’ performance on specific tasks. To solve the problem, you can instruct the model on the type of information you want, the structure or format by which you want it, etc. The process involves creating a prompt and an example response with the format and providing this info to the model. Next time you input to model the same type of prompt, it will give back a response in the way you desire, based on the example you provided.
In the abovementioned opportunities, you can use GAI to automate generic processes. But what if you want to automate tasks unique to your business? For example, you want a GAI model to generate technical documentation and user guides for your business products. A pre-trained language model cannot do this right away because it doesn’t know anything about your products. Fortunately, many large language models offer the option to fine-tune it.
Fine-tuning is the method of tuning a pre-trained model on a specific task to produce the desired results. In this example, the process involves tuning the pre-trained model with details about your business products. Fine-tuning LLMs with different types of data can enable you to leverage the power of GAI to automate tasks using that data. One example is that you can build a tool to automate sentiment analysis of new customer feedback by fine-tuning the model with your business’ customer feedback data. If a customer post negative feedback, GAI can detect that, and you can use the response from the model to call your CRM API and notify the concerned person to take the necessary steps. Likewise, you can think of many use cases across different departments; For example, sales intelligence automation to find the right salesman to engage a customer, automated evaluation of interview data, etc.
There are many different generative AI models available for different generative tasks. For example, there are numerous models for tasks like text generation, image generation, music generation, video generation, character generation, mathematical reasoning, scientific reasoning, etc. You can mix these models to expand the automation capabilities.
For example, a tool like ChatGPT can only provide you with textual responses. You can ask the model to suggest creative ideas for a poster, which will give you ideas as texts, not as an image. On the other side, tools like Stable Diffusion and Mid Journey will provide you with generative visual content based on your text inputs. Creating a workflow allows you to connect different models to automate complex tasks more efficiently.
The use case and applications are more than just content generation. You can use a fine-tuned model, as explained in the above opportunity, in the model pipeline. Now, this model pipeline can execute custom business tasks more efficiently. For example, imagine you provide your business’ inventory details as input to a fine-tuned text-based model. Now the model can process the prompt and generate a prompt as an output. The generated prompt can be provided as input to a mathematical modeling model to calculate the optimized inventory plan for the next month. Likewise, a combination of models can solve your business challenges or streamline your current business operations.
It will take a lot of work for a beginner to discover what problem can be solved using GAI models and how to do it optimally. The fruitful option would be to get an AI consultation from companies like Accubits with expertise in this area.
There are numerous Generative AI models available in the market. Many are open-source, meaning you can download and use them for free. The opportunity I want to explain here can be summarized in a phrase.
A Jarvis for your business
Of course, the phrase would only make sense if you know Jarvis, the Scifi character Iron Man’s AI assistant, that is the backbone of his powers. Imagine having an all-knowing AI assistant just for your business. You can use an open source language model and train the model with every information about your business. Or fine-tune a pre-trained model with department-specific information. Once done, you have a powerful AI assistant that can make a big difference in your organization. Interdepartmental communications and collaboration would become simple as eating a pie. An employee can ask the assistant about the leave policies; A CEO can get the sales forecast just by asking a bot rather than requesting and doing follow-ups with different people; A customer can ask about product details and offers to the same bot. In a nutshell, one bot to rule them all.
As previously said, this tech-driven change disrupting our world is inevitable and happening faster than ever before. Every day you can see the news of some organizations adopting Generative AI for different use cases and applications to modernize their business. LinkedIn has recently introduced an AI-powered writing suggestion feature in its platform, enabling users with automated content generation for LinkedIn profiles and helping recruiters write job descriptions. The feature is built on advanced GPT models. The company said;
“Our tool identifies the most important skills and experiences to highlight in your About and Headline sections and crafts suggestions to make your profile stand out. By doing the heavy lifting for you, the tool saves you time and energy while still maintaining your unique voice and style.”
PwC announced a partnership with AI startup Harvey to innovate in the Tax and Legal space. This announcement comes a month after the law firm Allen & Overy partnered with Harvey to integrate GAI models within its legal practice. Grammer assistant tool Grammarly plans to roll out the beta version of GrammarlyGo, a personalized GAI tool to complement its existing writing software. As said by the company, the tool understands personal writing and brand style in context, according to the company, and will generate relevant, comprehensive text on demand.
“Generative AI represents an inflection point in innovation that Grammarly can incorporate to deliver even more value for our customers,” said Rahul Roy-Chowdhury, Global Head of Product at Grammarly.
Salesforce made its move by launching Einstein GPT, a generative AI Customer relationship management technology, which delivers AI-created content across every sales, service, marketing, commerce, and IT process.
The best time to brainstorm on how your business can embrace AI was many years ago. The second best time is right now. Discovering how to leverage this opportunity could put your business WAY ahead of your competitors. As an expert in this field, we can help you navigate this global disruption, enabling you to lead the change. Have you got questions? Feel free to hit the ‘Ask Author’ button on the right side of this article. I’m happy to address your doubts and questions.
