text
stringlengths 0
182
|
---|
past technologies that have come and |
gonethink metaversethis latest one |
looks set to stay. OpenAIs chatbot, ChatGPT, is perhaps |
the best-known generative AI tool. It reached 100 million |
monthly active users in just two months after launch, |
surpassing even TikTok and Instagram in adoption speed, |
becoming the fastest-growing consumer application in |
history. |
1 |
Several innovations inherent within generative AI |
have obvious potential benefits for businesses. Its |
large language models (LLM) can learn from even |
bigger quantities of data than classical AI, including |
incorporating information from unstructured inputs. |
Users can interact with generative AI tools and receive |
responses in natural language. And finally, according to |
user needs, generative AI canas the name indicates |
generate a range of new outputs including text, |
pictures, computer code, and data streams. |
Such breakthroughs have led to high expectations. |
According to a McKinsey report, generative AI could |
add $2.6 trillion to $4.4 trillion annually in value to the |
global economy.2 The banking industry was highlighted |
as among sectors that could see the biggest impact (as |
a percentage of their revenues) from generative AI. The |
technology could deliver value equal to an additional |
$200 billion to $340 billion annually if the use cases |
were fully implemented, says the report. |
For businesses from every sector, the current challenge |
is to separate the hype that accompanies any new |
technology from the real and lasting value it may bring. |
This is a pressing issue for firms in financial services. |
The industrys already extensiveand growinguse |
of digital tools makes technology advances particularly |
likely to affect the sectors companies. |
According to a recent UBS report on the impact |
of generative AI, statistics from the U.S. Bureau of |
Labor suggest that banks and insurance are among |
the industries with the greatest proportion of their |
workforces exposed to potential automation.3 This |
MIT Technology Review Insights report delves further |
into that possibility, examining the early impact of |
generative AI within the financial sector, where it is |
starting to be applied, and the barriers that need |
to be overcome in the long run for its successful |
deployment. |
The following are the reports key learnings: |
Corporate deployment of generative AI in financial |
services is still largely nascent. The most active |
use cases revolve around cutting costs by freeing |
employees from low-value, repetitive work. Companies |
have begun deploying generative AI tools to automate |
time-consuming, tedious jobs, which previously required |
humans to assess unstructured information. Employees |
are thereby freed for more creative work, and in some |
cases, the tools outperform people. The following are |
common areas of deployment: |
5 |
MIT Technology Review Insights |
Customer service: Generative AI chatbots are |
already deployed to assist human customer service |
agents and may soon directly advise customers on basic |
questions. An academic study found that they increase |
customer service agent productivity and consumer |
happiness. |
Fraud prevention and risk management: The |
technology allows software to incorporate a wider and |
richer set of data into risk and fraud detection. |
Coding: Tools that can produce bespoke code for |
specific tasks have begun to appear. In a controlled |
experiment, an Australian bank found that these raised |
programmer productivity by 46%. |
Information analysis and summarization: |
Generative AI tools are already creating summaries of |
business conversationsa previously human task. On a |
larger scale, BloombergGPT provides the company |
another channel to monetize its archives by letting |
subscribers search them using natural language |
questions. |
There is extensive experimentation on potentially |
more disruptive tools, but signs of commercial |
deployment remain rare. Academics and banks are |
examining how generative AI could help in impactful |
areas including asset selection, improved simulations, |
and better understanding of asset correlation and tail |
riskthe probability that the asset performs far below |
or far above its average past performance. So far, |
however, a range of practical and regulatory challenges |
are impeding their commercial use. |
Legacy technology and talent shortages may slow |
adoption of generative AI tools, but only temporarily. |
Many financial services companies, especially |
large banks and insurers, still have substantial, |
aging information technology and data structures, |
potentially unfit for the use of modern applications. |
In recent years, however, the problem has eased with |