ChatGPT for Dummies

The last couple of months my LinkedIn feed has been overwhelmed with posts about the new kid on the block called chatGPT. This gives the feeling that this AI-tool is already a mainstream tool known by everyone.

Nonetheless when talking to friends and family less involved in IT or not working in the Tech sector, chatGTP appears still to be an unknown. Therefore I thought it might be interesting to give an introduction to chatGPT for dummies, so let’s start with the basics.

What is chatGPT?

ChatGPT (abbreviation for Chat Generative Pretrained Transformer) is a tool, based on a large AI-based language model, which allows to generate human-like text (i.e. indistinguishable from text written by humans).

The model, trained on millions of texts found on the internet (exact amount is not publicly known, but believed to be in the range of hundreds of billions of words), generates texts by adding each time the word which has the highest correlation with all previous words in the conversation. As such it can be used to generate any content (e.g. blogs, poems, market studies…), reply questions, but also execute more specific tasks, like e.g. generating programming code, translating texts, summarizing long documents, analyzing the sentiment of texts… As such it is generally recognized as the most advanced AI chatbot at the moment.

You can access it by going to https://chat.openai.com/ and registering. Afterwards you can converse with chatGPT for free (for now at least, it is likely that it will become paying in the future).

The success of chatGPT is enormous. In the first 5 days after its launch (on November 30, 2022) it counted already 1 million users (it took Facebook 10 months and Instagram 2,5 months to reach same numbers) and two months after its launch it reported to have 100 million users (it took TikTok 9 months and Instagram 2,5 years to reach this number of users).
It comes therefore as no surprise that Google initiated internally a code red, as they fear that chatGPT could be a danger for their flag product and core revenue source, i.e. the Google Search Engine.

Who is behind chatGPT?

ChatGPT is made available by OpenAI, which is together with DeepMind (part of the Google-Alphabet group and most famous for the AI model AlphaGo, which won in Go against the world champion) the most prominent player on the AI market at the moment.

OpenAI was founded in 2015 in San Francisco as a non-profit organisation to promote and research artificial intelligence (AI). The founders of the organisation are all famous Tech entrepreneurs like Elon Musk, Peter Thiel (Paypal co-founder and one the most known tech investors in Silicon Valley), Sam Altman (former president of the startup accelerator Y Combinator and now CEO of OpenAI), Jessica Livingston (founding partner of Y Combinator), Reid Hoffman (co-founder of LinkedIn)…
In 2019 it transitioned into a commercial (for profit) company, with Microsoft investing heavily in the company, i.e. in 2019 Microsoft already invested $1 billion in the company and in January 2023 Microsoft announced a multi-year investment of $10 billion. Additionally Microsoft is planning to integrate chatGPT in its search engine Bing, which could be a game-changer, allowing for Bing to break the dominant position of Google Search Engine.

ChatGPT is furthermore not the only successful AI model offered by OpenAI. E.g. in 2021, OpenAI also launched Dall-E, which is an AI tool that allows to generate any kind of digital image based on a natural language description.

Should I use it?

chatGPT is incredibly powerful and versatile, as it can generate any content in any style in a matter of minutes. As such it can be an enormous aid for increasing the productivity of any knowledge worker. While the quality, speed and efficiency of the output of chatGPT is amazing, it comes still with a few limitations as well. As such the content produced by chatGPT should not be copy/pasted without any reflection and revision.

Typical limitations of chatGPT are:

The model is only as good as the data it was trained on. As the model was trained on public internet pages, some training data can be biased, erroneous or of low-quality, which will reflect also in the answers of chatGPT. The more specific the topic of the prompt, the more limited the training data set will have been, meaning a higher chance for quality issues.
Especially the inherent bias of the model can be problematic and has already resulted in ethical debates. Numerous examples circulate on the internet where chatGPT showed strong bias and generated even discriminatory or even racist replies.
The generation of its output is very computationally intensive and requires significant resources. This can make it difficult or costly to use in certain settings, such as on low-power devices or in real-time applications. The CEO of OpenAI already mentionned in an interview that the cost per question is a few dollar cents. This seems low, but considering 100 million users executing several prompts each, you can do the math of how much money is being spent by OpenAI at the moment. You could argue that this money could be spent much better, as the majority of prompts are just for recreation and could be solved just as easily by a Google Search, which is much cheaper.
Due to its success, the system has become quite overloaded, which results in regular unavailability and general slowness for processing prompts. OpenAI has recently launched a paying service called ChatGPT Plus (at 20$ / month), which allows to get priority (i.e. access during peak times), faster responses and priority access to new features and improvements.
ChatGPT’s knowledge cutoff is September 2021, which means it is based on information that was available up until that date. It may not have knowledge of events or updates that have occurred since then.
Despite its incredible complexity, ChatGPT is still only a model that determines the best correlating word based on all previous conversation. This means the model does not have a deep understanding of the world or the ability to reason like a human. As a result some responses might be completely incorrect.

Some of those limitations are expected to be improved in the next version, i.e. GPTI 4, which is expected to be released in 2023.

In conclusion, yes you should use it, because it can help you enormously in improving your productivity, but remain sceptical of the result and use it only as a support tool.

How should I use it?

As mentioned before, after you are logged in, you can just ask any question and the reply will be formulated. After this reply, you can ask follow-up questions. In these follow-up questions, chatGPT will take into account the preceding conversation.

But the way you ask your question will also impact the quality of the response. The more precise your question the better the result and it’s advised to use certain guidelines (standards) in the question. Some people have already made it their profession to help companies providing the best prompts to chatGPT.

A good prompt typically consists of 3 parts:

First tell chatGPT who you are. This is done by a phrase like "Act like a …" or "You are a…" (e.g. "Act like a copywriting expert" or "You’re an expert career advisor"). Additionally you can provide more context allowing to really set the scene.
Then explain what you expect, i.e. define a task to complete (e.g. via "Your task is now to…"). Optionally you can provide additional specific context of your domain. This will help chatGPT to be more accurate.
Finally explain how. Often this is done by expressing the language, the style (e.g. "In the style of …") or the format (e.g. "In 5 bullet points") you want the reply to be in.

Some other good practices are:

Try to be as specific as possible in your prompts, i.e. the more specific the better the results will be.
Use correct language, i.e. avoid grammar mistakes, abbreviations or informal language
Ask to-the-point questions and avoid combining multiple questions in 1 prompt
As chatGPT keeps track of all previous context, ask follow-up questions on a reply, as those allow the model to provide answers with much more context and more tailered to your need.

You can find some ideas of prompts on https://www.awesomegptprompts.com/ or https://www.temok.com/blog/chatgpt-prompts/.

The most important is of course to try it out yourself and play with the system, until it gives you the desired result. There are no wrong prompts, but with some practice and the above tips, the results you will get will be much more valuable.

Note: some phrases in this blog were generated by chatGPT.
Note 2: the picture associated to this blog was generated by Dall-E.

Comments

Transforming the insurance sector to an Open API Ecosystem

1. Introduction "Open" has recently become a new buzzword in the financial services industry, i.e. open data, open APIs, Open Banking, Open Insurance …, but what does this new buzzword really mean? "Open" refers to the capability of companies to expose their services to the outside world, so that external partners or even competitors can use these services to bring added value to their customers. This trend is made possible by the technological evolution of open APIs (Application Programming Interfaces), which are the digital ports making this communication possible. Together companies, interconnected through open APIs, form a true API ecosystem , offering best-of-breed customer experience, by combining the digital services offered by multiple companies. In the technology sector this evolution has been ongoing for multiple years (think about the travelling sector, allowing you to book any hotel online). An excelle...

Are product silos in a bank inevitable?

Silo thinking is often frowned upon in the industry. It is often a synonym for bureaucratic processes and politics and in almost every article describing the threats of new innovative Fintech players on the banking industry, the strong bank product silos are put forward as one of the main blockages why incumbent banks are not able to (quickly) react to the changing customer expectations. Customers want solutions to their problems and do not want to be bothered about the internal organisation of their bank. Most banks are however organized by product domain (daily banking, investments and lending) and by customer segmentation (retail banking, private banking, SMEs and corporates). This division is reflected both at business and IT side and almost automatically leads to the creation of silos. It is however difficult to reorganize a bank without creating new silos or introducing other types of issues and inefficiencies. An organization is never ideal and needs to take a numbe...

RPA - The miracle solution for incumbent banks to bridge the automation gap with neo-banks?

Hypes and marketing buzz words are strongly present in the IT landscape. Often these are existing concepts, which have evolved technologically and are then renamed to a new term, as if it were a brand new technology or concept. If you want to understand and assess these new trends, it is important to reduce the concepts to their essence and compare them with existing technologies , e.g. Integration (middleware) software ensures that 2 separate applications or components can be integrated in an easy way. Of course, there is a huge evolution in the protocols, volumes of exchanged data, scalability, performance…, but in essence the problem remains the same. Nonetheless, there have been multiple terms for integration software such as ETL, ESB, EAI, SOA, Service Mesh… Data storage software ensures that data is stored in such a way that data is not lost and that there is some kind guaranteed consistency, maximum availability and scalability, easy retrieval...

IoT - Revolution or Evolution in the Financial Services Industry

1. The IoT hype We have all heard about the "Internet of Things" (IoT) as this revolutionary new technology, which will radically change our lives. But is it really such a revolution and will it really have an impact on the Financial Services Industry? To refresh our memory, the Internet of Things (IoT) refers to any object , which is able to collect data and communicate and share this information (like condition, geolocation…) over the internet . This communication will often occur between 2 objects (i.e. not involving any human), which is often referred to as Machine-to-Machine (M2M) communication. Well known examples are home thermostats, home security systems, fitness and health monitors, wearables… This all seems futuristic, but smartphones, tablets and smartwatches can also be considered as IoT devices. More importantly, beside these futuristic visions of IoT, the smartphone will most likely continue to be the cent...

PSD3: The Next Phase in Europe’s Payment Services Regulation

With the successful rollout of PSD2, the European Union (EU) continues to advance innovation in the payments domain through the anticipated introduction of the Payment Services Directive 3 (PSD3) . On June 28, 2023, the European Commission published a draft proposal for PSD3 and the Payment Services Regulation (PSR) . The finalized versions of this directive and associated regulation are expected to be available by late 2024, although some predictions suggest a more likely timeline of Q2 or Q3 2025. Given that member states are typically granted an 18-month transition period, PSD3 is expected to come into effect sometime in 2026. Notably, the Commission has introduced a regulation (PSR) alongside the PSD3 directive, ensuring more harmonization across member states as regulations are immediately effective and do not require national implementation, unlike directives. PSD3 shares the same objectives as PSD2, i.e. increasing competition in the payments landscape and en...

Trade-offs Are Inevitable in Software Delivery - Remember the CAP Theorem

In the world of financial services, the integrity of data systems is fundamentally reliant on non-functional requirements (NFRs) such as reliability and security. Despite their importance, NFRs often receive secondary consideration during project scoping, typically being reduced to a generic checklist aimed more at compliance than at genuine functionality. Regrettably, these initial NFRs are seldom met after delivery, which does not usually prevent deployment to production due to the vague and unrealistic nature of the original specifications. This common scenario results in significant end-user frustration as the system does not perform as expected, often being less stable or slower than anticipated. This situation underscores the need for better education on how to articulate and define NFRs , i.e. demanding only what is truly necessary and feasible within the given budget. Early and transparent discussions can lead to system architecture being tailored more close...

Low- and No-code platforms - Will IT developers soon be out of a job?

“ The future of coding is no coding at all ” - Chris Wanstrath (CEO at GitHub). Mid May I posted a blog on RPA (Robotic Process Automation - https://bankloch.blogspot.com/2020/05/rpa-miracle-solution-for-incumbent.html ) on how this technology, promises the world to companies. A very similar story is found with low- and no-code platforms, which also promise that business people, with limited to no knowledge of IT, can create complex business applications. These platforms originate , just as RPA tools, from the growing demand for IT developments , while IT cannot keep up with the available capacity. As a result, an enormous gap between IT teams and business demands is created, which is often filled by shadow-IT departments, which extend the IT workforce and create business tools in Excel, Access, WordPress… Unfortunately these tools built in shadow-IT departments arrive very soon at their limits, as they don’t support the required non-functional requirements (like h...

An overview of 1-year blogging

Last week I published my 60th post on my blog called Bankloch (a reference to "Banking" and my family name). The past year, I have published a blog on a weekly basis, providing my humble personal vision on the topics of Fintech, IT software delivery and mobility. This blogging has mainly been a personal enrichment , as it forced me to dive deep into a number of different topics, not only in researching for content, but also in trying to identify trends, innovations and patterns into these topics. Furthermore it allowed me to have several very interesting conversations and discussions with passionate colleagues in the financial industry and to get more insights into the wonderful world of blogging and more general of digital marketing, exploring subjects and tools like: Search Engine Optimization (SEO) LinkedIn post optimization Google Search Console Google AdWorks Google Blogger Thinker360 Finextra … Clearly it is not easy to get the necessary ...

The UPI Phenomenon: From Zero to 10 Billion

If there is one Indian innovation that has grabbed global headlines , it is undoubtedly the instant payment system UPI (Unified Payments Interface) . In August 2023, monthly UPI transactions exceeded an astounding 10 billion, marking a remarkable milestone for India’s payments ecosystem. No wonder that UPI has not only revolutionized transactions in India but has also gained international recognition for its remarkable growth. Launched in 2016 by the National Payments Corporation of India (NPCI) in collaboration with 21 member banks, UPI quickly became popular among consumers and businesses. In just a few years, it achieved remarkable milestones : By August 2023, UPI recorded an unprecedented 10.58 billion transactions , with an impressive 50% year-on-year growth. This volume represented approximately 190 billion euros . In July 2023, the UPI network connected 473 different banks . UPI is projected to achieve a staggering 1 ...

AI in Financial Services - A buzzword that is here to stay!

In a few of my most recent blogs I tried to demystify some of the buzzwords (like blockchain, Low- and No-Code platforms, RPA…), which are commonly used in the financial services industry. These buzzwords often entail interesting innovations, but contrary to their promise, they are not silver bullets solving any problem. Another such buzzword is AI (or also referred to as Machine Learning, Deep Learning, Enforced Learning… - the difference between those terms put aside). Again this term is also seriously hyped, creating unrealistic expectations, but contrary to many other buzzwords, this is something I truly believe will have a much larger impact on the financial services industry than many other buzzwords. This opinion is backed by a study of McKinsey and PWC indicating that 72% of company leaders consider that AI will be the most competitive advantage of the future and that this technology will be the most disruptive force in the decades to come. Deep Lea...

Bankloch

Search This Blog