Generative AI - Generally Intelligent or Generally Worrying?

According to McKinsey's recent report (The State of AI in 2023: Generative AI’s breakout year, 1 August 2023), generative AI could unlock a staggering $2.6-4.4 trillion in additional economic value annually.

This holds potential comparable to the GDP of major economies like the UK. Yet, companies and policymakers must tread cautiously, navigating risks around security, bias, jobs, and more to tap into these potential benefits. As Paul Daugherty, chief technology and innovation officer at Accenture, puts it, “The playing field is poised to become a lot more competitive, and businesses that don’t deploy AI and data to help them innovate in everything they do will be at a disadvantage.”

On the 18th of July 2023, an historic meeting at the UN Security Council initiated the first discussion about Artificial Intelligence and its global implications.

In an important address, James Cleverly, the Foreign Secretary, remarked,



Such predictions are echoed by Sundar Pichai, CEO of Alphabet, who suggested that “AI is one of the most important things humanity is working on. It is more profound than, I dunno, electricity or fire.” This year, the transformative nature of generative AI has really captured the world's imagination, with numerous startups and organisations bringing AI-powered products and technologies to market at an unprecedented pace. People no longer have to ask 'what is AI'? AI-generated text is now very convincing and human-like and is becoming a core part of the content creation and customer service functions of many companies.

Generative AI's rise heralds the dawn of a new era in automation. McKinsey's research suggests that the technology could automate a substantial 60-70% of tasks, up from the previously estimated 50%. This is largely due to advancements in natural language capabilities. As Pieter den Hamer, vice president of AI research at Gartner, states, “Every job will be impacted by AI. Most of that will be more augmentation rather than replacing workers.”

However, the tide of automation raises valid concerns. Andy Haldane, the former chief economist of the Bank of England, cautioned, "That hollowing out (of jobs) is going to be potentially on a much greater scale in the future when we have machines both thinking and doing - replacing both the cognitive and the technical skills of humans."



Still, former Slack CEO Stewart Butterfield offers a positive perspective, noting, “There's a lot of automation that can happen that isn't a replacement of humans but of mind-numbing behaviour.”
Of course, while AI may change the nature of many roles, it's unlikely to render the human touch obsolete. As Rob Thomas from IBM noted in 2020, “AI is not going to replace managers, but managers who use AI will replace the managers who do not.”

To ensure the safe and beneficial development of AI, Google, among other companies, has been a proponent of responsible AI use. As Pichai writes, "What matters, even more, is the race to build AI responsibly and make sure that as a society we get it right." Google's commitment to this can be seen in its launch of the Google for Startups Growth programme and the social innovation fund on AI. These initiatives aim to address Europe's pressing challenges using AI.

While the promises of generative AI are numerous, so are the risks. Bias, misinformation, security concerns, reliability, and intellectual property rights are just a few of the challenges that businesses and society at large will face. Transparent oversight, ethics review boards, and updated regulations are vital to ensure that this technology aligns with human values and standards.

Financial Services: A Closer Look

The financial services sector stands to be one of the biggest beneficiaries of generative AI, with McKinsey estimating it could unlock $200-340 billion in value annually. Banks and insurers have already been exploring uses like intelligent chatbots, predictive analytics and anti-fraud systems. But generative AI takes it to the next level with capabilities like:

Personalised marketing content and financial advice at scale

BNP Paribas is testing AI copywriters to localise and personalise marketing materials across regions and languages.

Hyper-targeted customer service

Chatbots like Bank of America's Erica can field common queries - it has already surpassed 1.5 billion interactions over 5 years and helped power a 19% rise in earnings -  and provide bespoke guidance for complex issues. Claude and Bard could take it further with more natural conversations.

Advanced fraud detection

Generative AI can model individual customer behaviour to spot anomalies that could point to the possibility of identity theft or suspicious transactions.

Improved risk modelling

Next-gen algorithms may better account for economic regime changes and black swan events compared to traditional models.

Operational efficiencies

JP Morgan is using AI to review commercial loan agreements much faster than lawyers. Contract analysis and other workflows could accelerate.

However, financial firms must be vigilant about reliability, bias and transparency as they integrate generative AI into high-impact functions. It seems clear though, that significant benefits will accrue to those companies who implement governance frameworks proactively rather than reactively.

Healthcare: Cautious Optimism

Healthcare is often cited among the top potential beneficiaries of AI, with McKinsey estimating a $300-590 billion value opportunity from generative AI techniques. However, significant risks around reliability and trust suggest a measured approach is sensible and pragmatic.

In fields like medical imaging, AI is already proving adept at spotting anomalies. Natural language models could streamline tasks like note-taking and report-writing, but patient treatment is an entirely different matter.

As the recent controversy around chatbot startup Anthropic shows, outputs cannot yet be taken at face value. Rigorous clinical validation is a must before considering any real-world application in areas like diagnosis or prescription.

Beyond technical robustness, generative AI must align with healthcare's guiding principle of "first, do no harm." Diversity and transparency in training data and algorithms will help avoid unfair treatment or exclusion of underrepresented groups. Strict ethics governance can ensure AI complements rather than compromises patient agency and the human touch.

For now, the greatest promise resides in more bounded applications like accelerating drug discovery through computational trials of millions of molecular combinations. With a measured approach, generative AI could unlock lifesaving insights while upholding healthcare's exacting standards. But irresponsible implementation risks setbacks to adoption and public trust.

Retail: New Frontiers in Personalisation

Preparing for the Generative AI Era