Sometimes you can look back at something and say – that was a turning point. It doesn’t feel it when you are living through it, and it can take years or decades to realise that it was a seminal change to how things operated from there on out. One fairly recent example would be the Gmail account. Prior to the launch of Gmail in 2004, a person’s email account was limited to their computer and had a finite amount of space attached to it. Hotmail, Yahoo mail, etc introduced web based email accounts that could be accessed from any internet café, but they were limited in size. When Google launched the Gmail account they went big – every user was given 1GB of storage space and with Google being Google, it became easier than ever to not only store your emails, but search them too. Threads of emails could be grouped into conversations that could be accessed from any device anywhere in the world. We now have access to our emails on multiple devices at any time, and can search from users to key words within your inbox.
You may have had some interactions with Chatbots when you’re trying to find information on the web – a little pop up box that asks whether you need help when you’re browsing a store, or when you’re trying to get a tax certificate or the like from a financial institution. The JSE calls their chatbot Jesse … cute but not particularly bright yet! The release of ChatGPT, the chatbot interface powered by the most advanced Artificial Intelligence (AI) algorithms and available for free to anyone with access to the internet, feels like a giant leap from what we’re used to and another one of those seminal moments.
Artificial Intelligence is not new. Self-driving cars, intelligent sensors used in agriculture and even good old google search algorithms are all examples of very niche uses of artificial intelligence. ChatGPT feels different in that it’s the closest we’ve got to Artificial General Intelligence (AGI). Whether you’re a chef looking for some ideas for a new recipe, or an investor looking for some deeper technical research into a set of stocks, you can pose any question and get a decent, intelligent response. Not only that, but with ChatGPT, it becomes a conversation rather than a refinement of search terms.
Browsing on Google already involves a lot of artificial intelligence – the algorithms that drive your search in one way or another, and rank the entries from their various databases to what is most relevant to you. The search function is not interactive though. You can refine your search, choose different key words, but at ultimately the search engine is only interrogating a database and returning results.
ChatGPT was developed by a company called OpenAI and was launched to the public late in 2022 after it had finished it’s “training”. The language is conversational which allows you to ask follow up questions, and go back and explore finer details. I am not nearly technical enough to know how it works, but I know that this is machine learning in progress and the more that it is used, and the more data that is available to it, the more refined it will become.
Think about some of the real world applications where it would be useful:
- in customer service systems, where it could accurately respond to everyday language;
- to translate text between languages which would breakdown language barriers in business;
- to create reports and articles where content is required;
- to assist healthcare professionals with diagnostic capabilities; and
- to analyse financial data to provide investment decisions and risk management.
Interestingly, I asked ChatGPT a few questions about a tight chest and a lingering cough and although it diagnosed a dose of flu, it did advise me to seek confirmation from a medical professional. I also asked a few questions about asset allocation and a bit about the current economic environment and it counselled me a few to times to consult a financial advisor!! So hopefully we humans will have a role to play at some level!
There are some ethical questions that need to be considered too. Large amounts of personal data will be collected and processed, so it will be important that this data will be used in an ethical and responsible manner. AI models can amplify biases in society if they are trained on biased data. It is quite scary to imagine the impact of these biases. It will be difficult to determine who is responsible for the actions and decisions made by AI systems and in a world where accountability is lacking, it would become almost impossible to hold anyone accountable if real world consequences were guided by artificial intelligence. And of course, the spreading of misinformation or “fake news” could be used maliciously if controls are not in place.
Policy makers, parents and educators will need to prioritise media literacy – especially for younger generations – in order to provide them the tools to make informed decisions about the information they consume. When I think about what real people can achieve through manipulation of data – think Cambridge Analytica and the election rigging scandal – it does petrify me that a clever machine could be far more manipulative. Unfortunately I don’t think we can dial back the technological advances that we are seeing, so the advances come with a responsibility to ensure that they are used in a positive and ethical manner.
As an example, I asked ChatGPT the following question: I am a financial advisor and am writing a comment on the use of AI and all the above issues – could you help me write the comment? This is the response:
Sure, here’s a sample monthly comment on the use of AI in finance and the related ethical considerations:
As financial advisors, we are always looking for ways to provide our clients with the best possible investment advice. With the rapid advancements in artificial intelligence, we are seeing increasing use of AI in the financial sector. While AI has the potential to bring significant benefits to the industry, it is also important to consider the ethical considerations that come with its use.
One key issue is the potential for bias in AI systems. If the algorithms used in AI systems are trained on biased data, they may produce biased results, leading to discriminatory outcomes and negatively impacting marginalized communities. It is important to ensure that the data used to train AI systems is diverse and free from bias to minimize these risks.
Another important consideration is privacy. With the increasing use of AI in finance, large amounts of personal data may be collected, stored, and processed. Ensuring the protection of personal data and using it in an ethical and responsible manner is critical.
In addition, it is important to determine who is responsible for the actions and decisions made by AI systems in finance. As AI systems become more sophisticated and make increasingly complex decisions, accountability and transparency will become even more important.
Finally, we must also be mindful of the potential job displacement that may result from the increasing use of AI in finance. It is important to address this issue through policy and re-training programs to ensure that individuals are equipped with the skills needed for the jobs of the future.In conclusion, while AI has the potential to bring significant benefits to the financial sector, it is important to consider the ethical considerations that come with its use. By being aware of these considerations and taking a responsible approach, we can ensure that AI is used in a manner that is beneficial for all.
Interestingly, Elon Musk (who was a co-founder of OpenAI) has said that artificial intelligence is one of the “greatest threats to humanity” and has called for government regulation to oversee the rollout of AI. It may slow the rampant use of AI, which isn’t necessarily a bad thing. For a man who is not a fan of regulation and oversight, it makes his call all the more meaningful. It definitely offers food for thought!! You can read more on his thoughts by following this link: Yahoo Finance Article
The rise of Artificial Intelligence
Sometimes you can look back at something and say – that was a turning point. It doesn’t feel it when you are living through it, and it can take years or decades to realise that it was a seminal change to how things operated from there on out. One fairly recent example would be the Gmail account. Prior to the launch of Gmail in 2004, a person’s email account was limited to their computer and had a finite amount of space attached to it. Hotmail, Yahoo mail, etc introduced web based email accounts that could be accessed from any internet café, but they were limited in size. When Google launched the Gmail account they went big – every user was given 1GB of storage space and with Google being Google, it became easier than ever to not only store your emails, but search them too. Threads of emails could be grouped into conversations that could be accessed from any device anywhere in the world. We now have access to our emails on multiple devices at any time, and can search from users to key words within your inbox.
You may have had some interactions with Chatbots when you’re trying to find information on the web – a little pop up box that asks whether you need help when you’re browsing a store, or when you’re trying to get a tax certificate or the like from a financial institution. The JSE calls their chatbot Jesse … cute but not particularly bright yet! The release of ChatGPT, the chatbot interface powered by the most advanced Artificial Intelligence (AI) algorithms and available for free to anyone with access to the internet, feels like a giant leap from what we’re used to and another one of those seminal moments.
Artificial Intelligence is not new. Self-driving cars, intelligent sensors used in agriculture and even good old google search algorithms are all examples of very niche uses of artificial intelligence. ChatGPT feels different in that it’s the closest we’ve got to Artificial General Intelligence (AGI). Whether you’re a chef looking for some ideas for a new recipe, or an investor looking for some deeper technical research into a set of stocks, you can pose any question and get a decent, intelligent response. Not only that, but with ChatGPT, it becomes a conversation rather than a refinement of search terms.
Browsing on Google already involves a lot of artificial intelligence – the algorithms that drive your search in one way or another, and rank the entries from their various databases to what is most relevant to you. The search function is not interactive though. You can refine your search, choose different key words, but at ultimately the search engine is only interrogating a database and returning results.
ChatGPT was developed by a company called OpenAI and was launched to the public late in 2022 after it had finished it’s “training”. The language is conversational which allows you to ask follow up questions, and go back and explore finer details. I am not nearly technical enough to know how it works, but I know that this is machine learning in progress and the more that it is used, and the more data that is available to it, the more refined it will become.
Think about some of the real world applications where it would be useful:
Interestingly, I asked ChatGPT a few questions about a tight chest and a lingering cough and although it diagnosed a dose of flu, it did advise me to seek confirmation from a medical professional. I also asked a few questions about asset allocation and a bit about the current economic environment and it counselled me a few to times to consult a financial advisor!! So hopefully we humans will have a role to play at some level!
There are some ethical questions that need to be considered too. Large amounts of personal data will be collected and processed, so it will be important that this data will be used in an ethical and responsible manner. AI models can amplify biases in society if they are trained on biased data. It is quite scary to imagine the impact of these biases. It will be difficult to determine who is responsible for the actions and decisions made by AI systems and in a world where accountability is lacking, it would become almost impossible to hold anyone accountable if real world consequences were guided by artificial intelligence. And of course, the spreading of misinformation or “fake news” could be used maliciously if controls are not in place.
Policy makers, parents and educators will need to prioritise media literacy – especially for younger generations – in order to provide them the tools to make informed decisions about the information they consume. When I think about what real people can achieve through manipulation of data – think Cambridge Analytica and the election rigging scandal – it does petrify me that a clever machine could be far more manipulative. Unfortunately I don’t think we can dial back the technological advances that we are seeing, so the advances come with a responsibility to ensure that they are used in a positive and ethical manner.
As an example, I asked ChatGPT the following question: I am a financial advisor and am writing a comment on the use of AI and all the above issues – could you help me write the comment? This is the response:
Interestingly, Elon Musk (who was a co-founder of OpenAI) has said that artificial intelligence is one of the “greatest threats to humanity” and has called for government regulation to oversee the rollout of AI. It may slow the rampant use of AI, which isn’t necessarily a bad thing. For a man who is not a fan of regulation and oversight, it makes his call all the more meaningful. It definitely offers food for thought!! You can read more on his thoughts by following this link: Yahoo Finance Article
Asset Class Returns
The table below represents a rolling year view of the major asset class returns that we track. It offers a view of the asset classes we use to diversify your portfolio.