A new report has found that 8% of decision-makers in the UK and 7% in the US plan to spend more than $25m (£19.5m) on AI initiatives this year, after a number of investors expressed concerns about the technology’s capabilities and potential return on investment.
According to the State of AI 2024 report by consultancy Searce, a quarter of decision-makers said their organisations intend to spend between $11m (£8.5m) and $25m (£19.5m) on AI in 2024.
The most common reason for making these investments was to stimulate new business development, as indicated by 31% of respondents in the UK and 35% in the US. From their perspective, the idea seems to be working.
More than 90% of UK decision-makers surveyed consider their AI initiatives to be “successful”, and almost a third plan to increase AI spending by up to 50%, with a further 8% saying they expect investment to increase by 100% or more.
The findings are based on a survey of 300 chief executives and senior technology executives at organisations with revenues of at least $500 million (£390 million) conducted in June and July this year.
The report finds that 70% of business decision makers already have at least three generative AI use cases that are ready to go and are operational with their investments. These include tools for customer service, internal research, content generation, marketing and sales, coding, data analysis, and capture.
SEE: Generative AI: UK business leaders face investment challenges as everyone claims to be an expert
However, while 51% of respondents considered their investments to be “very successful”, only 42% said they were “somewhat successful”.
“The difference between highly and moderately successful initiatives indicates a gap in AI implementation maturity,” the report’s authors write.
“This is an opportunity for organizations to focus on talent development, data quality, and solid evaluation metrics to enhance AI capabilities and achieve greater ROI.
Paul Pallath, vice president of applied AI at Searce, said in the report, “When it comes to AI, organizations often focus on short-term, easy-to-achieve goals, which results in significant technology and process debt that becomes costly to manage as organizations grow.”
The authors added: “To achieve true return on investment (ROI), organisations need to move away from mindlessly investing in these initiatives and hoping for the best, and instead adopt an outcomes-focused approach, underpinned by appropriate governance, measurable change management frameworks and processes.”
UK business leaders concerned about lack of AI talent to support their investments
Respondents were asked about the AI implementation challenges they were most concerned about. For UK decision-makers, it was the lack of skilled talent, cited by 19%.
The level of ‘skill gap vacancies’ – where a job cannot be filled because candidates lack the skills, qualifications or experience – is very high in the UK ICT sector, rising from an already high 25% in 2017 to 43% in 2022, the latest year for which data is available.
The latest report also found that the UK is the 25th most tech-savvy country in Europe, well behind other digital leaders in the region such as Germany, France and Spain.
SEE: Key UK skills gaps include artificial intelligence and strategic thinking, says Red Hat
The UK government has recognised the digital skills shortage in the country and has made a number of key investments over the last year or more to try to address this. In November 2023, it was announced that over £200 million would be allocated to support colleges and universities to provide more training opportunities in industries including digital.
In March this year, Science Secretary Michelle Donelan unveiled a further funding package worth more than £1.1 billion to fund 4,000 PhDs in engineering and physical sciences.
Microsoft has also made significant investments in bridging the digital skills gap in the UK. In December 2023, the tech giant announced a “multi-million pound investment” to provide AI skills training to more than one million people.
US Business Leaders Concerned About Privacy and Security of Their AI Data
According to the Searce study, the biggest barrier to AI adoption was another barrier for U.S. decision makers, as 20% of respondents said it was data privacy and security. This is linked to concerns about protecting confidential information, constantly changing regulations and maintaining customer trust.
“This skepticism may stem from high-profile failures or failure to meet lofty expectations, leading to caution in AI investments,” the authors wrote. Such high-profile failures include DPD’s chatbot cursing customers, Microsoft’s Twitter bot trolling, and Google’s Bard (now Gemini) model answering a question incorrectly during its reveal, causing Alphabet to lose $100 billion in shares.
There is also tangible evidence that AI’s capabilities are being overestimated. A recent Stanford study found that AI is still not as good as humans at complex tasks of advanced mathematical problem-solving, visual common-sense reasoning, and planning.
A 2022 Deloitte report found that the percentage of organizations in the “weak” AI category (large adoption, poor performance) rose from 17% to 22% over the year, suggesting that performance is weak.
SEE: Data Scientist Survey: Do Tech Leaders Believe the AI Hype?
Investors are concerned about the return on AI investments — and companies should be too
Julian Mulhare, Managing Director, EMEA, Searce, said in a press release accompanying the report: “As global AI investment continues to grow, as our research shows, it is critical for companies to focus not just on spending, but also on the tangible benefits these investments can deliver.”
Investors are highlighting this fact, expressing concerns about when, or if, massive cash injections into tech companies developing AI models will bear fruit.
Jim Covello, a Goldman Sachs analyst, wrote in a June report: “Despite its high price tag, the technology is still far from where it needs to be to be useful… Overbuilding things the world has no use for or is not ready for usually ends badly.”
Sequoia Capital partner David Cahn argued in a blog post that the AI industry would need to generate a whopping $600 billion per year to cover hardware costs.
SEE: UK tech startups see first decline of 2022, down 11% this quarter
According to S&P Global, combined capital spending by Microsoft, Alphabet and Meta increased 60% year over year as a result of AI investments. Alphabet alone spent $13 billion in the second quarter, 91% more than in the second quarter of 2023, putting pressure on profit margins.
But Meta CEO Mark Zuckerberg noted on its second-quarter 2024 earnings conference call that he expects it will be “years” before the company monetizes its AI products. Less-than-reassuring comments like Zuckerberg’s contributed to shares of the “Magnificent Seven” U.S. tech companies—NVIDIA, Meta, Alphabet, Microsoft, Amazon, Tesla, and Apple—losing a combined $1.3 trillion over a five-day period in early August.
What’s more, while business leaders may be excited about the prospect of increased internal efficiency thanks to AI, consumers don’t necessarily share that optimism. A new study from Washington State University found that the presence of the term “artificial intelligence” in a product description actually “reduces purchase intent.”
This is largely due to a lack of trust in AI capabilities and perceived risks associated with elements such as loss of control and privacy. Therefore, companies should be aware of the ROI of using AI in their products as well as in internal implementations.