Hire or AI?
Every manager, especially in the financial services industry, must be asking the question – do I hire new staff, or do I leverage AI? How do I use AI when everything I have is already working and what are my peers doing?
Being over 2 years since OpenAi opened the floodgates for normal users to leverage AI tools using large language models, there is now a groundswell of interest in learning about how AI is going to help my business. Quant funds and large financial institutions have been using data based decisions for trading, risk management or scenario assessment for a while now, however, small and large firms are grappling with the question on how to use AI in a meaningful way.
To further confuse the thought process, there are now large language models as well as small language models (those that require less compute power and perform limited “specific” tasks), there is agentic AI (AI that does multiple tasks in sequence, acting as an agent of the instructor - whether that is a portfolio manager, research assistant or lawyer), in addition to the plethora of AI tools like NLP, machine learning, deep learning, inference learning. Many of these technologies depend on the transformer invented by Google Brain which essentially processes all words in parallel using a mechanism called the self-attention mechanism, which allowed models to process entire sequences of text in parallel, making them much faster and more efficient than previous architectures like recurrent neural networks (RNNs) and long short-term memory (LSTM) networks.
There are also many providers of technology – the cloud providers like Microsoft, Amazon, Google, Oracle, IBM etc. the large language models besides Chat GPT 4, Anthropic, Llama 3, Mistral etc. Most vendors including those like Snowflake, DataBricks, Salesforce, ServiceNow, Microsoft, Meta have introduced the concept of co-pilot, a smart colleague, but then AI as the person. In addition, there are many industry verticals and specific products that leverage some form of AI, whether it is machine learning, deep learning, large language models, or now agentic AI. These companies have specific solutions for targeted industries and use cases and include vendors like Robin.ai, Alphastream.ai, Untap.ai, Nomad Data, Canoe Intelligence, Accelex, Jarvis, Harvey.ai, BlueFlame.ai, Alkami.io amongst others.
So, if I am a manager, CTO, or business head, how do I proceed? Based upon many dialogues with managers in the financial industry, we offer some structure on best practices to help you select and implement an AI System.
Utilize AI solutions that are industry specific, and in particular, use case specific. Besides the general utility tools like Microsoft Copilot, or Chat GPT, we recommend a specific use case that you want to “pilot” your first AI technology.
Do not re-invent the wheel – leverage a third-party vendor solution. There are many vendors who have spent lots of time thinking of use cases, have expertise in data analytics and AI tools and have come up with solutions. As you build more expertise, you will have opportunities to come up with proprietary IP specific for your business model.
The cost of failure is too high. Many projects in software and third-party vendor implementation do not succeed. So, be specific and chose carefully, ideally partner with a consulting firm, to help you select and implement a specific AI solution. If the first project fails, there will be corporate skepticism which will delay the implementation of other solutions, putting you at risk and competitive disadvantage to your peers and competitors.
Think of the initial implementations as strategic band-aids that may last 1 – 3 years. Many of the AI solutions are evolving and are not very expensive – especially if they hold true to their promise of huge efficiencies. The rate of change is rapid so you may have new products and solutions that you may incorporate in the future.
Do not wait – the cost of waiting is too high. Unlike other eras where early adopters or bleeding edge tried out new things and some failed and some succeeded, AI is different in that it is real and can make a dramatic change to a use case or to a business.
In summary, this is truly an exciting time for businesses – as they can get lots of work done without hiring lots of employees. There is a true revolution going on that will drastically change how we perform work and how we live. The early (ier) adopters will gain the “intelligence” muscle and enjoy advantages over those who don’t. Even if we go back to the 10,000 hours or 10 years benchmark to build expertise, the sooner you start, the sooner you build your intelligence, in using artificial (computer) intelligence.
About the Author:
Jayesh Punater is a successful Entrepreneur, Investor, Fintech Strategist and Thought Leader. He focuses on bringing innovative technology and business models to the FinTech industry. Mr. Punater is the Founder and CEO of Nucleus DNA, a FinTech studio and venture investment platform. Nucleus invests in, advises, and provides services to start-up and advanced stage scale-up firms. Mr. Punater also co-founded Nucleus EMV – a venture capital fund that invests in LATAM financial technology firms. Jayesh Punater is a successful Entrepreneur, Investor, Fintech Strategist and Thought Leader. He focuses on bringing innovative technology and business models to the FinTech industry. Mr. Punater is the Founder and CEO of Nucleus DNA, a FinTech studio and venture investment platform. Nucleus invests in, advises, and provides services to start-up and advanced stage scale-up firms. Mr. Punater also co-founded Nucleus EMV – a venture capital fund that invests in LATAM financial technology firms.
