Enterprises are rolling out more AI – to 'middling results'
- Reference: 1667208669
- News link: https://www.theregister.co.uk/2022/10/31/deloitte_ai_enterprise/
- Source link:
The 5th Edition of Deloitte's [1]State of AI in the Enterprise report is based on a survey of 2,620 business leaders from organizations around the globe, all of whom are responsible for AI technology spending or managing its implementation.
According to the authors, the AI race (if such a thing ever existed) is no longer about adopting AI or automating processes for efficiency, but has now moved on to realizing value, driving outcomes, and unleashing the potential of AI to drive new opportunities.
[2]
However, the topline findings are that many organizations are struggling with "middling results" despite increased deployment activity since the last edition of the report.
[3]
[4]
According to Deloitte, 79 percent of respondents claimed to have achieved full-scale deployment of three or more types of AI applications, up from 62 percent last year. But also up was the percentage of those rating their organizations as "underachievers" – 22 percent in this report compared with 17 percent last time.
Underachievers are characterized by Deloitte as organizations that have carried out a significant amount of development and deployment activity yet failed to achieve the outcomes they were looking for.
[5]
In spite of this, 76 percent of respondents reported that their organization plans to increase its investments in AI "to gain more benefits." This figure is slightly down from the 85 percent that planned to increase investment last year, indicating that funding may be leveling off after the last few years of significant increase. Only 3 percent of respondents reported a decrease in investment.
According to Deloitte, organizations cited different challenges depending on what stage they are at in their AI project implementation. Justifying the business value is the number one reported challenge when starting new AI projects, perhaps not surprisingly.
However, once organizations attempt to scale up their AI projects, other impediments to progress come to the fore, such as managing AI-related risks, lack of executive buy-in, and lack of maintenance or ongoing support.
[6]Shutterstock partners with OpenAI to sell AI-generated stock images using DALL-E
[7]Human-replacing AI startups reach $1bn unicorn status
[8]Microsoft said to be in talks to invest more in OpenAI
[9]AI programming assistants mean rethinking computer science education
"This demonstrates the ongoing challenge of establishing the coordination and discipline needed to consistently fund initiatives after they have ceased to be the shiny object," the authors state, adding that building an "AI-fueled organization" requires discipline and focus to maintain the resulting systems and algorithms so that they will continue to generate value.
For those respondents that are further along the road of adoption, 87 percent reported that they now find the length of payback time for AI projects comes within their expectations or sooner.
[10]
But Deloitte cautions that while this may indicate an increased understanding of implementation issues, it might also suggest that organizations are too focused on AI projects for cost savings instead of the "transformational opportunities" that AI can offer.
In fact, reduced costs were reported by 78 percent of respondents as the most desired outcome, leading the report authors to warn that more "transformational outcomes," such as revenue generation or business innovation, may be being ignored.
Deloitte makes the point that leadership and culture are important for organizations seeking to successfully deploy AI. It said it found that high-outcome organizations from its survey were over 55 percent more likely to invest in change management compared to low-outcome organizations.
However, only 43 percent of respondents said that they had appointed a leader responsible for effective human and AI collaboration, and only 21 percent reported actively educating workers on when to apply AI most effectively.
But perhaps the key message of the report is that orgs have to redesign their business operations around AI if they want to get the full benefit of it. This sounds a little backwards to our mind – surely technology is supposed to adapt to the way we work, not the other way round.
The report states that despite evidence that establishing processes and redefining roles to deliver better quality AI results in improved outcomes, there has been little growth in the market in terms of adopting such practices. ®
Get our [11]Tech Resources
[1] https://www2.deloitte.com/us/en/pages/consulting/articles/state-of-ai-2022.html
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2Y1@q4v@fOduUNS3UtehmXQAAANE&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y1@q4v@fOduUNS3UtehmXQAAANE&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Y1@q4v@fOduUNS3UtehmXQAAANE&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y1@q4v@fOduUNS3UtehmXQAAANE&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[6] https://www.theregister.com/2022/10/26/shutterstock_getty_openai_dall_e/
[7] https://www.theregister.com/2022/10/23/two_generative_ai_startups_javis/
[8] https://www.theregister.com/2022/10/21/microsoft_openai_investment/
[9] https://www.theregister.com/2022/10/20/ai_programming_tools_mean_rethinking/
[10] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Y1@q4v@fOduUNS3UtehmXQAAANE&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[11] https://whitepapers.theregister.com/
1) Collecting data is expensive.
2) Classifying data, well, requires human eye to "train" data, usually in the form of an expert on that data set
3) Errors have to be cleaned up to remove outliers in your dataset
4) Weak, historical database designs, ancient paper records that won't go away, and/or spreadsheet reliance complicate the job of relating one piece of information to another. The role of an AI programme isn't to redesign your data hierarchy; and attempting to do so will "break" various other business functions without a very bottom up re-write of "everything".
5) Employing 20 people to develop an expert system to "replace" the job of an actual expert on subject matter in question? Vastly more expensive.
and last, but not least
6) Having done all that farce with your AI box, what are you going to do with the output? (Relates to lack of experts).
I speak from a position of "not for want of trying". I have ran several R&D programmes over the years to try and find something useful to do with AI in the absence of large teams of experts to draw on. The only ones that consistently come up with anything usable are in the image recognition space.
Totally agree. I've worked in both public and private sectors where "AI" makes for exciting and promotion-inducing PowerPoint slides for executives, but the less sexy slides of the underlying data infrastructure to enable it are skipped over entirely, leaving data scientists with the 2 circles and then the owl meme when their group fails to deliver. The cultural shift to funding and creating machine-readable, re-useable data across organisations remains the greatest challenge of "industry 4.0", and in many cases would add value in itself just by enabling department heads to see it for the first time.
Recursive AI
...that many organizations are struggling with "middling results"...
They should put the success/fail modalities in an AI and predict whether they will succeed or fail. Then they have a successful AI to predict their downfall?
Middling?
From what I've seen it's more muddling. In several meanings of the term. Sometimes concurrently.