More Than Chatbots: AI Trends Driving Conversational Experiences For Customers

An important benefit of using Google Gemini is that its supporting knowledge base is as large as any chatbot’s—it’s created and updated by Google. So if your team is looking to brainstorm ideas or check an existing plan against a huge database, the Gemini app can be very useful due to its deep and constantly ChatGPT updated reservoir of data. In a growing trend across the AI chatbot sector, the Crisp Chatbot can be customized to match a business’s branding and tone. This is increasingly important in crowded markets where a number of companies are seeking to create a distinct brand to cut through the clutter.

It discusses exploratory data analysis, regression approaches, and model validation with tools such as XLMiner. The training is appropriate for anybody interested in using generative vs conversational ai data to acquire insights and make better business decisions. A $49 monthly Coursera subscription gives you access to the lecture materials as well as a certificate.

Speech-to-text technology lies at the very core of conversational intelligence – everything else comes from there. Second, we also see a rise in smaller (and cheaper) generative AI models, trained on specific data and deployed locally to reduce costs and optimise efficiency. Even OpenAI, which has led the race for ever-larger models, has released the GPT-4o Mini model to reduce costs and improve performance. This widely used model describes a recurring process in which the initial success of a technology leads to inflated public expectations that eventually fail to be realised. After the early “peak of inflated expectations” comes a “trough of disillusionment”, followed by a “slope of enlightenment” which eventually reaches a “plateau of productivity”. The AI assistant is customisable to accomplish specific tasks to ensure the application of industry policies.

Multimodality and LLMs

There, a technician tasked with making sure a customer-facing bot can understand and respond to customers appropriately is able to use LLMs to auto-generate new and more appropriate training data for the bot. Also, while Alexa has been integrated with thousands of third-party devices and services, it turns out that LLMs are not terribly good at handling such integrations. Overall, the former employees paint a picture of a company desperately behind its Big Tech rivals Google, Microsoft, and Meta in the race to launch AI chatbots and agents, and floundering in its efforts to catch up.

Moreover, the technology will reduce false positives by 30% in application security testing and threat detection by 2027. In healthcare, the usage of generative AI is creating new ways of enhancing patient care and accelerating research activities. Deloitte’s 2024 Life Sciences and Health Care Generative AI Outlook Survey reveals that 75% of healthcare companies are experimenting with this technology.

Related content

The integration of ChatGPT in teaching and learning can significantly impact educators’ roles and the entire teaching-learning process. ChatGPT can revolutionize traditional instructional practices with its interactive and conversational capabilities and open new possibilities for personalized and engaging learning experiences. ChatGPT faces several challenges that must be addressed to improve its performance and ethical considerations. Language models are trained on vast amounts of text data, which may inadvertently contain tendencies in the data sources. Addressing biases requires careful data curation, identification, and mitigation techniques to ensure fairness and inclusivity in the AI model’s responses.

Meta builds technologies that help people connect, find communities, and grow businesses. Apps like Messenger, Instagram, and WhatsApp further empowered billions around the world. Now, Meta is moving beyond 2D screens toward immersive experiences like augmented and virtual reality to help build the next evolution in social technology. Historically, AI in ecommerce has been around for quite some time — just not in the way we have seen it leveraged lately. For example, Google has built its business on search algorithms that are essentially artificial intelligence. When you shop online, that storefront personalizes products and recommendations based on who you are, what you like and what you’ve previously purchased.

Early signs of success are already evident, with 15 million SMBs using WhatsApp for Business to create digital presence and drive traffic through click-to-chat ads. With the advent of generative AI-powered assistants and ease of integration with conversational platforms, these conversational journeys can now be implemented at scale with much faster deployment cycles. This is driven by the capabilities of generative AI assistants, enabling contextualized, humanlike conversations with reasoning ability, multimodal support, and vernacular language proficiency. The investments by leading tech players to democratize access to generative AI platforms and cultivate an ecosystem of offerings will further fuel this new era of consumer engagement. Implementing chat-based assisted journeys, known as conversational journeys, on platforms with high user engagement (e.g., social media and messaging) can be key for businesses to engage and facilitate online transactions. This is already in motion—most consumers are informally engaging with both small and large businesses (e.g., messaging carpenters, doctors, bank representatives, and direct-to-consumer brands) on social media and messaging platforms.

The user interface (UI) for machine learning applications typically involves dashboards and visualizations that display analytical results, predictions, and trends. These interfaces are designed to help users interpret data insights and make informed decisions. In contrast, generative AI interfaces often include tools for content creation, such as text editors, image generators, and design software.

The company’s solutions are custom tailored to each business, and can be customized without the need for extensive coding. Specializing in the development of generative AI solutions for the contact center, Cresta gives companies solutions that pinpoint the drivers of performance, sales, and customer conversations. With AI-native copilots, QA, and coaching solutions, trained on company data, organizations can discover new ways to increase revenue, and customer satisfaction scores. Calabrio’s speech analytics solution turns raw conversational data into usable customer intelligence, with predictive net promoter scores, sentiment indicators, and automated agent evaluations. Calabrio also pairs conversational data with meta data stripped from screen recordings and keyboard activities, for full end-to-end visibility.

Automating Monotonous Tasks

Generative AI tools tend to come in the form of chatbots, powered by large language models (LLMs). LLMs apply this deep learning to vast data sets to understand, summarize, and generate new content. Genesys Cloud CX is an all-in-one, AI‑powered cloud contact center solution that enables organizations to personalize end-to-end experiences at scale. It has a built-in Agent Assist tool with an auto-summarization functionality that creates instant summaries of customer conversations. You can foun additiona information about ai customer service and artificial intelligence and NLP. The solution also integrates predictive analytics and natural language processing (NLP) to understand customer sentiment and intent, refining personalization of customer engagements. Last but not the least, Genesys Cloud CX has an open API framework that lets organizations incorporate additional GenAI solutions to modify the platform to their specific needs.

For example, if a user wrote that he was feeling angry because he got in a fight with his mom, the system would classify this response as a relationship problem. Later in Woebot’s development, the AI team replaced regexes with classifiers trained with supervised learning. The process for creating AI classifiers that comply with regulatory standards was involved—each classifier required months of effort. Typically, a team of internal-data labelers and content creators reviewed examples of user messages (with all personally identifiable information stripped out) taken from a specific point in the conversation.

The “Analyze” offering forms part of the comprehensive “Eureka” platform from CallMiner, combining deep AI analysis with automated journey mapping, automatic interaction scores, and even predicted NPS scores. There are also robust APIs available to connect your customer insights to your CRM, Business Intelligence tools, and other data repositories. CallMiner also offers secure automatic redaction, customizable reports, and organization-wide alerting. According to Ranger, you can do that with a well-built conversational AI chatbot or voice. “We have seen a huge demand now for the traditional sort of conversational AI products to solve a specific problem within a business. We are seeing that in retail, but universally across the board and in retail, it is mostly around customer service, post-sales,” he noted.

We reviewed each AI chatbot pricing model and available plans, plus the availability of a free trial to test out the platform. The platform is a web-based environment allowing users to experiment with different OpenAI models, including GPT-4, GPT-3.5 Turbo, and others. OpenAI Playground is suitable for advanced users looking for a customizable generative AI chatbot model that they can fine-tune to suit their business needs. This advanced platform enables a vast level of choices and approaches in an AI chatbot. Replika is an artificial intelligence chatbot designed to have meaningful and empathetic-seeming conversations with users.

Part of a comprehensive suite of intelligent cloud tools offered by Google, DialogFlow is a solution for building conversational agents. The system leverages the vendor’s resources for generative AI and machine learning, providing a single development platform for both chatbots and voice bots. AI company Aisera produces a wide suite of products for employee, customer, voice, Ops, and bring-your-own-bot experiences.

The goal is to boost SoundHound Chat AI’s ability to answer questions about events almost as they happen and make SoundHound’s voice assistant more useful in vehicles and other devices. It aimed to provide for more natural language queries, rather than keywords, for search. It also had a share-conversation function and a double-check function that helped users fact-check generated results. Generative AI is transforming industries by enabling the use of powerful machine learning models to create new content. As the need for AI-powered solutions grows, understanding generative AI may lead to new opportunities, both personally and professionally.

Ultimately, our research intends to support creative and student-centered teaching and learning techniques while facilitating the successful integration of ChatGPT into education. Stakeholders may make intelligent decisions about ChatGPT’s deployment and use it to improve educational experiences by knowing its benefits, challenges, and ethical issues. We do not just discuss biases, outdated data, transparency, and legitimacy; we work to fix them. Our research also focuses on the ethical side, ensuring data privacy, inclusivity, and a good balance between AI and human interaction.

Rapid innovation cycles driven by GenAI will enable banks to stay ahead of the curve and effectively cater to evolving customer demands. Such capabilities of LLMs – such as GPT, PaLM and Falcon – have led to deployments of conversational AI skyrocketing across numerous industries and all stages of the customer journey. “Most of the GPUs are still A100, not H100,” the former Alexa LLM research scientist added, referring to the most powerful GPU Nvidia currently has available. None of the hyperscalers or other GenAI app providers offer customers an end-to-end capability to experiment with a range of LLM or SLM models to develop, deploy, and manage sophisticated GenAI apps. In doing so, they can choose from 30+ LLMs, including community, open-source, and finetuned models. Moreover, the vendor allows users to apply different models to different apps to optimize their performance.

Ensuring that the GenAI systems comply with such industry regulations as GDPR, CCPA, or HIPAA is imperative to avoid legal ramifications. Kore.ai’s latest CX Benchmark report highlights that UK consumers are comfortable with using AI in their banking interactions and would be happy having more AI Automated Assistants supporting them. “Lack of mature technology, adequate policies and procedures, training, and safeguards are creating a perfect storm for AI accidents far more dramatic than just hallucinations. Yet, as these businesses begin to dream bigger with their use of GenAI, there is much more to consider.

So AI companies are still at work on bigger and more expensive models, and tech companies such as Microsoft and Apple are betting on returns from their existing investments in generative AI. According to one recent estimate, generative AI will need to produce US$600 billion in annual revenue to justify current investments – and this figure is likely to grow to US$1 trillion in the coming years. For example, generative AI systems can solve some highly complex university admission tests yet fail very simple tasks. This makes it very hard to judge the potential of these technologies, which leads to false confidence. Privacy and security measures provided by Einstein Trust Layer protect information from unauthorised access and data breaches through zero-data retention from Salesforce’s LLM partners.

As AI advances, agent assist tools can generate personalized responses to customer queries in seconds, track sentiment scores, and streamline onboarding processes. Therefore, Sallam (2023) has systematically analyzed the prospective views and legitimate concerns regarding using ChatGPT in healthcare education. The author thoroughly analyzes ChatGPT’s application in healthcare education, considering both optimistic perspectives and legitimate concerns. Based on a comprehensive analysis of 70 research publications, the author investigates the utility of large language models in healthcare teaching, research, and practice. According to the author, ChatGPT’s promising uses could lead to paradigm shifts in medical practice, study, and training.

Because of AI tools, businesses can now expand content production without compromising quality. AI-driven technologies such as ChatGPT have the potential to increase ChatGPT App productivity and streamline tedious administrative activities. The increase in AI and human interaction will be primarily facilitated by deep learning algorithms.

We built technical safeguards into the experimental Woebot to ensure that it wouldn’t say anything to users that was distressing or counter to the process. First, we used what engineers consider “best in class” LLMs that are less likely to produce hallucinations or offensive language. Finally, we wrapped users’ statements in our own careful prompts to elicit appropriate responses from the LLM, which Woebot would then convey to users.

• It stands out for its ability to understand and generate human-like responses, making it an effective tool for customer support, personal assistance, and general information retrieval.
• We were excited by the possibilities, because ChatGPT could carry on fluid and complex conversations about millions of topics, far more than we could ever include in a decision tree.
• Companies can use the conversational analysis tools offered by IBM to build data fabrics, predict outcomes in interactions, and customize customer care.
• And our newest community, VKTR, is home for AI practitioners and forward thinking leaders focused on the business of enterprise AI.
• We already know that no matter how many you contract or hire, they’re already fully utilized by the time they walk in on their first day.

He is passionate about using math and software to improve lives, and has used his senior leadership positions at tech companies including Samasource and Alt12 Apps to help reduce poverty in Africa and improve women’s health. He holds three bachelor’s degrees from MIT in mathematics, philosophy, and management science. The AI team faced the question of whether LLMs like ChatGPT could be used to meet Woebot’s design goals and enhance users’ experiences, putting them on a path to better mental health. With ChatGPT, conversations about mental health ended quickly and did not allow a user to engage in the psychological processes of change.

• So I think that’s what we’re driving for.And even though I gave a use case there as a consumer, you can see how that applies in the employee experience as well.
• Privacy and data protection should be paramount when deploying ChatGPT in an educational setting.
• Plus, companies can also use intelligent insights to analyse employee performance, and identify specific skill and knowledge gaps that could be limiting growth.
• In truth, these two AI apps are highly distinct, which should make your choice an easy one.

This can then be used to help with agent training or to provide notes and suggestions during the call to steer the conversation and keep the customer satisfied. Large language models also display so-called emergent abilities, which are unexpected abilities in tasks for which they haven’t been trained. Researchers have reported new capabilities “emerging” when models reach a specific critical “breakthrough” size. Research shows that the size of language models (number of parameters), as well as the amount of data and computing power used for training all contribute to improved model performance.

For instance, users can choose a persuasive or creative writing mode to tailor the AI’s assistance to their needs. The learning curve for implementing machine learning solutions is generally steep, so you’ll need a solid understanding of statistics, data science, and algorithm development. You may also need to be proficient in data preprocessing, model training, and evaluation. CX automation company Verint offers conversational AI solutions in the form of its chatbots, IVA, and live chat toolkit. With this ecosystem, businesses can build comprehensive conversational workflows with bots that support digital, SMS, voice, and mobile channels. Verint Voice and Digital Containment bots use NLU and AI to automate interactions with all types of customers.

Customers in the U.S., the UK, and India have already asked Rufus tens of millions of questions, and we’re excited to introduce it in these countries too. “Oracle does not want its APEX platform take over all application development, serving as a general-purpose, low-code development environment for any and all use cases,” said Bradley Shimmin, chief analyst at Omdia. The AI Assistant also can be used to add new pages, edit existing pages, or add security features to the application, the company said. “The assistant can help a developer identify errors in the SQL code and also explain the next steps required to fix the code,” the senior vice president said. This menu, according to Hichwa, is aimed at helping developers iterate and refine SQL queries.

While they perform distinct functions, both technologies are interrelated and frequently complement one another. By implementing an AI-powered virtual assistant powered by IBM® watsonx Assistant™, the organization has dramatically increased both responsiveness and customer satisfaction. The assistant, named Trinny, interacts with website visitors in real time, fielding 120 frequently asked questions in natural language. Making numerous strides in the world of generative AI and conversational AI solutions, Microsoft empowers companies with their Azure AI platform. The solution enables business leaders to create intelligent apps at scale with open-source models that integrate with existing tools.

3 Ways I Use Copilot to Improve My Microsoft Excel Experience

Nearly 100,000 Karya workers are recording voice samples, transcribing audio or checking the accuracy of AI-generated sentences in their native languages, earning nearly 20 times India’s minimum wage for their work. Karya also provides royalties to all contributors each time its datasets are sold to AI developers. The diverse ecosystem of NLP tools and libraries allows data scientists to tackle a wide range of language processing challenges. From basic text analysis to advanced language generation, these tools enable the development of applications that can understand and respond to human language. With continued advancements in NLP, the future holds even more powerful tools, enhancing the capabilities of data scientists in creating smarter, language-aware applications. It’s available as part of the Nvidia AI Enterprise software platform, which gives businesses access to additional resources, including technical support and enterprise-grade security, to streamline AI development for production environments.

By integrating Copilot into Excel, Microsoft has given itself a competitive edge in both the desktop and online environments. In this article, I’ll break down a few of my favorite things Copilot can do in Excel and what I’d like to see it tackle in the future. Globally, the CCDH noted, some regulators have the power to investigate the claims in the CCDH’s report, including the European Commission under the Digital Services Act and the UK’s Ofcom under the Online Safety Act. X and the CCDH have long clashed, with X unsuccessfully suing to seemingly silence the CCDH’s reporting on hate speech on X, which X claimed caused tens of millions in advertising losses. During that legal battle, the CCDH called Musk a “thin-skinned tyrant” who could not tolerate independent research on his platform. And a federal judge agreed that X was clearly suing to “punish” and censor the CCDH, dismissing X’s lawsuit last March.

Explained: How AI Can Fight Sex And Gender Bias In Healthcare

Some data might be best kept encrypted in order to deny access from a proscribed country, which also would include Russia. If a customer asks for the latest news about a company, for instance, the system queries recent news documents. On the other hand, if the question is about stock performance, the model ChatGPT App accesses structured financial data to provide the current stock price and trends. The ability to reason about which tool to call upon demonstrates the system’s agentic capabilities. We support CTOs, CIOs and other technology leaders in managing business critical issues both for today and in the future.

For example, while AI can help with data storytelling, finance professionals need to communicate the insights that AI produces and turn them into a coherent narrative. You can’t wave it over shaky data and expect it to generate valuable insights or fix your data analysis problems. The adage of “garbage in, garbage out” applies to AI data analytics just as much as to manual analysis, and Gartner notes that poor data quality is often cited as a primary reason for slow AI adoption among finance teams. Companies that run troll farms also use data mining and large data sets obtained from users to predict and influence voters. One famous example is Cambridge Analytica, the data analytics firm that worked with Donald Trump’s election team and the winning Brexit campaign.

NLTK (Natural Language Toolkit)

After gaining initial access, the group conducts reconnaissance and may deploy additional malware such as the CSharp-Streamer remote access trojan. The Internet Archive’s crawl back to normalcy hit a bump when a threat actor sent many of the digital library’s users a spoofed email after apparently stealing a stolen access token for the site’s Zendesk account. Payment card services giant Visa said Wednesday that fraudsters are reverting to older methods, including credit card theft to quickly use stolen card information for gift cards, goods or online transactions. Encourage them to verify AI findings at first, to help them grow comfortable with working with the outcomes. Set specific advantages that you expect to gain from introducing AI, together with KPIs and metrics that you’ll track to measure success.

You can choose to focus on Excel specifically, as well as certain tasks and jobs, or just browse the entire collection of prompts. The benefit of conversational chatbots is they’re designed to understand standard questions. In theory, you should be able to ask Copilot questions about your dataset as you would ask an expert who know all about that data, and receive informed answers back. For example, you can ask Copilot to split your data across multiple columns by any metric you specify, perhaps if you want to split full names in your spreadsheet from one column into two (one for first names and one for last names).

It can analyze past interactions with candidates and predict which individuals may be a good fit for future openings, creating a more proactive approach to recruitment. For example, AI can analyze facial expressions, tone of voice, and speech patterns during interviews to assess a candidate’s level of engagement, emotional intelligence, or honesty. Bengaluru-based startup CoRover.ai already has over a billion users of its LLM-based conversational AI platform, which includes text, audio and video-based agents. I am extremely bullish about the future of Microsoft’s CoPilot because, unlike some of the other A.I. Chatbots out there, I can clearly see the productivity angle that Microsoft has emphasized with Copilot. Its integration with the Microsoft ecosystemincluding Windows itselfwill probably make it increasingly useful over time.

You can foun additiona information about ai customer service and artificial intelligence and NLP. Troll farms are “primarily a hybrid warfare tool, but in more democratic regimes, political parties also use such tools, and I’m certain it’s not entirely foreign in Romania,” Septimius Pirvu added. Subsequently, the president of the electoral authority AEP, Toni Greblă, denied competency and specified that he had already contacted other institutions involved in combating such practices, including the Ministry of Interior, the Digitalization Authority, and SRI. The head of the Permanent Electoral Authority added that such suspicions are a national security issue. Lasconi further emphasized that any involvement of foreign operatives in Romanian elections threatens national security and transparency.

While there are many potentially suspect ways to utilize the technology to cut corners and outsource work, there are also legitimate use cases for AI in your professional day to day. In fact, companies like Microsoft are now injecting AI tools directly into their work apps, in the hopes that users will find ways to make the most of the technology to enhance their workflows. Critics say the draft convention allows for intrusive domestic and cross-border surveillance with minimal limitations, risking misuse against dissenting voices. Provisions for collecting electronic evidence and international cooperation could facilitate human rights violations and conflict with existing EU data protection laws, signatories said.

• Lasconi published images of Tal Hanan, a businessman known for attempting to manipulate elections in 30 countries, in Bucharest at the Aspen Institute headquarters.
• To support initiatives like these, Nvidia has released a small language model for Hindi, India’s most prevalent language with over half a billion speakers.
• In addition to the 100,000 developers trained in AI in India, Nvidia said there have been an additional 100,000 academic and student developers trained as well.
• There are many free resources to help you learn and understand data structures and algorithms, which allow effective data processing and problem-solving in AI models.
• You could also ask the assistant to bold certain metrics that change as your data changes.
• On a platform with an estimated 429 million daily active users worldwide, only about 400 notes were displayed within the past two weeks in less than an hour of a post going live.

In 2018, it was revealed that the company harvested millions of Facebook profiles of US voters, in one of the tech giant’s biggest ever data breaches, and used them to build a powerful software program to influence elections. EWeek has the latest technology news and analysis, buying guides, and product reviews for IT professionals and technology buyers. EWeek stays on the cutting edge of technology news and IT trends through interviews and expert analysis. Gain insight from top innovators and thought leaders in the fields of IT, business, enterprise software, startups, and more. AI has the potential to revolutionise the recruitment industry by making the hiring process faster, more efficient, and more data-driven.

Serving over a billion local language speakers with LLMs

With that, let’s take a look at some of the Copilot features I think might be of use to Excel users. Spreadsheets are not my thing, so I imagine the following could offer support, especially for those of us that might not know exactly what we’re dataset for chatbot doing when we open up an overflowing page of numbers and figures. If you don’t want to pay for both, a Copilot Pro subscription does give you access to Copilot in the web versions of Excel, which Microsoft offers for free for everyone.

For notes that took longer—which the CCDH suggested is the majority if the fact-check is on a controversial topic—only about 60 more notes were displayed in more than an hour. Currently, more than 800,000 X users contribute to Community Notes, and with the lightning notes update, X can calculate their scores more quickly. That efficiency, X said, will either spike the amount of content removals or reduce sharing of false ChatGPT or misleading posts. This appears to be a common pattern on X, the CCDH suggested, and Musk is seemingly a multiplier. In July, the CCDH reported that Musk’s misleading posts about the 2024 election in particular were viewed more than a billion times without any notes ever added. In a report, the CCDH flagged 283 misleading X posts fueling election disinformation spread this year that never displayed a Community Note.

Risk Management Framework: Learn from NIST

However, it can’t interface with every data set that a user might want, which leaves you collecting data manually before feeding it into Excel. Data that the market intelligence firm Sensor Tower recently shared with Ars offers a potential clue as to why the CCDH is seeing so many accurate notes that are never voted as “helpful.” That’s the question the Center for Countering Digital Hate (CCDH) is asking after digging through a million notes in a public X dataset to find out how many misleading claims spreading widely on X about the US election weren’t quickly fact-checked. Fortunately, adjusting for such biases appears to lead to better healthcare outcomes for women. More studies are needed on this topic, but researchers have already proposed building AI-designed PPE. Ensuring that sex traits are considered in PPE design could be expected to improve safety.

• He also noted that troll accounts are easier to spot on some social media platforms such as Facebook and harder on others, like TikTok.
• This ensures, for example, that HR chatbots provide responses based on access rights, preventing unauthorised disclosures.
• “Governance remains a crucial aspect of AI adoption, with organisations establishing AI oversight boards and rigorously testing models before deploying them in production,” he said.
• Adoption is high, with a recent NVIDIA survey reporting that 91 per cent of financial service companies are either assessing or actively using AI to automate tasks and improve operational efficiency.
• The benefit of conversational chatbots is they’re designed to understand standard questions.

With rising demand for data-driven insights, the global decision intelligence industry is forecast to grow to $64 billion by 2034, up from $12.1 billion this year, according to Future Market Insights, Inc. Many finance teams hesitate to embrace AI solutions out of concerns that they could undermine data privacy or weaken data security. Data security is important, as handling large amounts of sensitive information requires robust protection measures. These concerns are well-founded, too – last year, Samsung banned employees from using third-party GenAI tools after ChatGPT leaked sensitive data. Artificial intelligence (AI) is blazing through finance departments like wildfire, particularly for financial planning and analysis (FP&A) use cases.

You can highlight specific data types, like if you want to see your top six most productive months highlighted in green, and your lowest productive months highlighted in red. You could also ask the assistant to bold certain metrics that change as your data changes. Rather than spend time trying to bold items yourself, or figure out how to spread data through multiple columns, see if a Copilot prompt can achieve the same result. I am amazed by Excel power users who intuitively know how to use formulas to perform calculations in their spreadsheets. I could probably fill up a table with data if need be, but ask me to come up with formulas to perform calculations in my dataset?

International regulations are also catching up with AI and establishing requirements around data privacy and security. It’s important to build clear policies around data use, set up and regularly review access permissions, and establish logging and monitoring to track unauthorised use or data access. Strategic decision-making is another area that needs to remain human-led, and finance personnel are needed to manage relationships with stakeholders in other departments. Compliance and ethics are areas that are growing in importance as AI becomes the norm and that should remain under human management. AI is still a shiny new object, but it’s a mistake to rush in blindly and adopt every AI tool you see.