Generative AI will be great for Generative AI consultants

Generative AI will be great for Generative AI consultants

Receive free AI updates

Here’s an unusual, sales note on generative AI that does more than just gobble and regurgitate hype. It comes from JPMorgan Tien-tsin Huang analysts et alwho deal with the bank’s IT services.

GenAI represents the largest technology wave since cloud or mobile, so it will be a multi-faceted revenue driver for IT services and BPO [business process outsourcing] suppliers, they write. For the next few years, however, the bulk of spending will be on organizing messy databases, disabuse management of disappointments, and direct investment in turbo-suitable businesses with what is effectively an unbridled form of autocomplete.

JPMorgan sees no shortage of applications for GenAI. But to demonstrate the limitations compared to the average wage slave, he highlights how quickly the probability-weighted algorithm powering ChatGPT gets distracted or bored whenever the job has fixed parameters:

A 13-word sentence where only 12 words begin with the letter L

GenAI is error prone. It’s not artificial general intelligence, or AI indistinguishable from human intelligence. It can’t code perfectly, and it can’t at all create a real corporate IT architecture, just like it can’t perfectly value a company, and it can’t create a real equity portfolio at all. For established companies with substantial business momentum, critical information systems, and customer relationships to maintain, responsiveness is critical. GenAI is a phenomenal productivity enhancer, but it’s not good enough to replace most of its users.

That’s how it’s worked so far, with generative AI taking over the wingman role for software developers, digital artists, and content mill workers. At the corporate level, the projects that are currently being worked on are mostly about building walled garden AI tools that are not at risk of data leakage and will likely be forever confined to the corporate intranet.

It’s likely that early public-facing implementations (aside from toys like Midjourney and ChatGPT) will aim to reduce customer service bottlenecks; things like call centers, where companies have built large language models with their own proprietary data, JPMorgan says.

By 2027 or so, more conservative businesses will be willing to take the lead from early movers so that AI can expand dramatically and exponentially, says JPMorgan. That’s when things like hyper-personalized media and robot PAs become a mainstream reality.

But for most businesses, the prep work will be the difficult and expensive part. The average enterprise IT network is simply too much of a clusterflux to use as a feedstock for generative AI, particularly when no one can predict or even explain what they will do with the data.

Here is JPMorgans’ conclusion in full:

Despite significant automation, engagement, and other opportunities, true widespread adoption of genAI remains years away. Early enthusiasts passed the baton to innovators in the typical adoption cycle, but the mainstream market has serious hurdles to overcome to truly capitalize on the opportunity genAI presents.

Business leaders’ first questions for their technology consultants when interacting with genAI tools are (1) how can we use it in our organization and (2) what do we need to do to implement it. Every question is critical. Jack Dorsey, co-founder and Block Head of Block, encouraged companies to approach investing in genAI from a use case perspective rather than a technology perspective in his comments at our 2023 TMC conference in Boston, suggesting that the excitement around technology could lead companies to spend aimlessly and therefore realize lower returns. IT service companies can help operating companies direct their genAI spend to the highest-performing areas. AC extension [Accenture] cited in its F3Q23 earnings call that the company completed 100 genAI projects amounting to ~$100 million in sales in the prior four months; While this initial traction demonstrates a strong positioning for ACN and reflects the firm’s strong customer relationships, that $100 million is still a minuscule figure based on ACN’s $60+ billion in revenue, and most of these engagements likely represent preliminary and exploratory projects as customers begin to understand how they want to use genAI. These exploratory engagements may include conversations about what type of model to use, how to train it, quantify the increment required to prepare training data, etc.

Legacy companies must accelerate their digital transformations with an emphasis on data preparation to reap the full potential benefits of genAI. Companies should take full advantage of genAI models when opening and refining them with their proprietary data; to do this, companies need to have their data in relatively good shape, which for most companies continues to be a challenge. Unified data neatly assembled in a modern cloud-hosted database is great for LLM setups, but most companies’ data is far from unified across anything but the modern database infrastructure. The tech debt we discuss at length when describing the drivers of digital transformation spending at legacy enterprises complicates genAI implementation just as it complicates other technology initiatives. Companies with bad data hosted in disparate legacy databases will benefit from consolidating and modernizing their data holdings in their efforts to implement genAI. The work required to prepare business data for model tuning will likely reduce model tuning work. The long-standing supply-demand gap for engineering talent (which persists even after supply easing due to cost rationalization of unprofitable and large-cap technology) ensures that firms are looking to technology services firms for assistance with this data cleansing, migration and unification job.

It took mobile about eight years to approach saturation, and cloud is still nowhere near saturation after more than a decade of massive investment. From ACN to GDYN [Grid Dynamics], the largest and smallest IT service providers in our coverage (and beyond) say cloud work remains responsible for the majority of their revenue growth. Tech debt is so large in most legacy institutions that any new technology faces a very steep climb. Therefore, despite our enthusiasm for genAI’s potential to drive profitability across our coverage, we expect it will be a few years for genAI to contribute significant cost savings and revenue growth for our coverage group, and many more years for these contributions to reach their full potential.

#Generative #great #Generative #consultants
Image Source :

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *