How often lately have you found yourself wondering, "Was this generated by artificial intelligence?" "This" could be anything—from a company growth strategy to a birthday greeting, in any format—text, graphics, images, voice, or music. I think it’s quite often. And just as often, we ask ourselves, "Is this a good thing or a bad thing?"
I don’t like to rely on subjective opinions. That’s why I’ve gathered the most recent and reliable data (such as those by McKinsey and Deloitte) to evaluate both the benefits and downsides of AI in the workplace.
Benefits of AI in the workspace
The Deloitte report "State of Ethics and Trust in Technology" has been one of the most exciting studies for me in the field of AI for business in 2024. Here are some key numbers from it:
- 94% of respondents said they are using GenAI, and 87% are increasing its use.
- 54% identified GenAI as the most significant risk among emerging technologies.
- The application of GenAI in high-tech sectors is expected to generate up to $460 billion, with the greatest impact in software engineering, marketing, and sales.
By the way, McKinsey estimated the contribution of GenAI to the global economy at $2.6 to $4.3 trillion per year.
Here are the top benefits of implementing AI in the workplace (source: Deloitte).
Credits: Deloitte
№1 Improved efficiency, productivity, revenue, and cost reduction.
AI has the potential to automate tasks, improve decision-making, and provide valuable insights into business operations. Businesses are using AI to improve efficiency and productivity by automating repetitive tasks and streamlining processes. For example, AI-powered chatbots can provide immediate and personalized responses to customer inquiries, freeing up human customer service representatives to handle more complex issues.
№2 Іncreased innovation, improved products, and services.
AI can help businesses develop new ideas and products. More companies experimenting with and implementing AI will likely lead to more innovation. Generative AI can create new content, such as images, text, music, and other media, that mimic human creativity. This capability can be used to develop new products and services.
№3 Enhanced user experience.
AI can revolutionize how you interact with customers, and if done right, it can make communication faster and more personalized. I’m talking about things that leading companies are already implementing—personalized messaging, tailored recommendations, and customized user flows—all of which create a unique and impressive user experience that ultimately boosts conversions and profitability.
Just as ASOS and Microsoft did in August 2024 when they launched an AI-powered experience “to surprise and delight young fashion lovers”.
Credits: Lefty
To define an overall AI strategy, as well as address specific challenges in the workplace, I recommend answering three questions.
1. Which business functions can AI be effective for?
Experience from other companies shows that AI integration delivers the most value in customer relations, marketing, software development, and R&D. At Sommo, we’ve had the chance to test AI in each of these areas (across different projects). We conclude that even if a business function is critically important, it doesn’t mean it can’t be enhanced with AI (or vice versa). However, it’s crucial to allocate sufficient time and effort for its development and testing.
2. Which tasks or processes within these functions can be automated or enhanced with AI?
This can be achieved by continuously analyzing and testing AI tools within the context of these tasks and processes. The number, quality, and functionality of AI tools are constantly evolving. That’s why I recommend making an analysis of AI capabilities a part of strategic planning.
3. What KPIs will we use to evaluate AI implementation?
Assessing effectiveness is always the most challenging and essential part of any strategy. We didn't reinvent the wheel and just tracked key metrics for each business function/task:
- Customer operations: problem resolution speed, ticket processing time, customer satisfaction (NPS).
- Marketing and sales: number of leads, conversion rate, engagement/retention, churn rate, sales volume.
- Development: coding speed, number of errors.
- R&D: research and design time, modeling and testing accuracy, and new product development speed.
At Sommo, we actively leverage GenAI to accelerate product development, speed up time to market, and capture a larger audience. From writing code to generating ideas for user retention and creating AI support bots, GenAI is an integral part of our process. Our excitement about GenAI's success always comes with understanding how fleeting this success can be and the need for constant oversight. Roman and I discussed this in more detail, with examples, in episode 2 of our podcast.
What is the negative of AI in the workplace?
Anyone who has used AI tools knows that it can be risky. Sometimes it's just amusing, but there are plenty of cases where real harm is done. I’ve grouped the main risks into three categories to help eliminate them more effectively:
1. Data security and privacy. To provide relevant answers, an AI model needs to know as much as possible about you (your company, the task, your client). Establishing clear, open rules for data use and ensuring compliance at all levels of the company is absolutely essential.
2. Reliability and accuracy. So-called "hallucinations" (inaccurate or fabricated AI information) have become a frequent meme. One part of building a culture of AI use should be mandatory quality checks and continuous monitoring.
3. Bias. This is not just about ethical and reputational risks, but also about discriminatory outcomes and even potential harm.
Credits: Deloitte
Here are some questions you should ask to estimate your company's costs of Generative AI.
What data do we have available that could be used to train and fine-tune Generative AI models, and how can we ensure data quality and security? Assess your data's quality, availability, and security to ensure it is suitable for training GenAI models. Consider factors such as data privacy, bias, and compliance with regulations.
What are the potential risks associated with implementing Generative AI in these areas, and how can we mitigate them?
What investments in technology, infrastructure, and talent will be required to implement and scale Generative AI effectively? Determine the resources needed for successful implementation, including hardware, software, cloud computing, and skilled personnel.
How can AI boost productivity/efficiency in this workplace? Implementing AI should never be an end in itself, no matter how interesting, exciting, or trendy it may seem.
How can the value created by Generative AI be measured and demonstrated to stakeholders, such as employees, clients, and investors?
How can we foster a culture of responsible AI usage within our company, ensuring ethical considerations are prioritized throughout the development and deployment process? Establish clear guidelines, training programs, and governance structures to promote responsible AI practices.
Getting more of AI in the workspace
In my opinion, a perfect analogy for the implementation of AI in the workspace sounds like this:
"An LLM is like an engine. No one just wants the engine of a car or a plane; they want a car or a plane. So, there are all these things you need to do to make it part of business processes so the business can use it."
Like other technologies, generative AI will only realize its full potential when integrated into everyday tasks. While early projects have shown promising results and driven increased investment in generative AI, organizations must demonstrate consistent and significant value as quickly as possible.
Credits: Deloitte
Consider the following actions/principles to get more benefits and reduce the risks associated with AI in your company's workspace.
- Write an AI usage policy.
- Implement regular evaluations of AI tools.
- Regulate data processing.
- Complement AI with human involvement.
- Communicate and address issues openly.
- Encourage initiatives, experiments, and pilot projects.
- Measure and evaluate AI implementation results transparently.
In the next article, we’ll dive into how to create an AI acceptable use policy, complete with a handy template to guide you through the process. Stay tuned!
