Artificial Intelligence – is it bad or is it good?
The flames of fear that Artificial Intelligence could wipe out 300 million jobs over the next few years were only fanned last month when a group of movers and shakers in the world of AI development suggested that the technology could spell the end of mankind.
An armageddon prophecy, perhaps, but similar predictions of doom and gloom were made in the early 19th century on the dawn of the Industrial Revolution when a group of angry and fearful English weavers called the Luddites broke into the new steam powered textile factories and smashed the mechanised looms over the same fear that their livelihoods would be destroyed.
We know what happened as a consequence of the Industrial Revolution and how, despite the terrible Dickensian factory conditions, it eventually led to the creation of a middle class and the modern capitalist and consumer systems we know it today. The term ‘luddite’ has therefore passed into the English vernacular as meaning someone who stubbornly resists the inevitable tide of change, even though in the long-term it will be beneficial to society and mankind.
So are we in Portugal behaving like luddites in fearing for the future, rather than embracing this new AI technology (actually, not so new since its first steps can be traced back to the 1950s) which has the potential to allow the human race to make huge leaps in terms of technology in all areas?
This question was debated at a special event organised by the American Chamber of Commerce in Portugal (AmCham) on Friday at a debate at the Sheraton Hotel ‘AI – Hype or Game Changer’ at which it was decided that as with all sudden technological leaps, there is a quite a bit of both.
The panel of experts invited to debate the issue were: Professor Arlindo Oliveira, an illustrious academic and author, Daniela Braga, founder and CEO of Portuguese DNA unicorn Defined.ai, Pedro Marques, Head of Advanced Analytics and Intelligence (Coordinator) at SPMS, EPE – Ministry of Health Shared Services, Madalena Talone, Executive Board and Commission director for public bank Caixa Geral de Depósitos (CGD), with the debate moderated by Manuel Maria Correia, Managing Director of DXC and President of AmCham’s Special Commission for Technology and Innovation. The event was introduced by Pedro Duarte, Corporate, External & Legal Affairs Director at Microsoft Portugal.
Pedro Duarte, calling AI the “hot topic of the moment” that six months ago people knew about but weren’t really talking about, said that everyone is now discussing how AI can be regulated and controlled, which spoke volumes about “how far this technology is advancing”.
Pedro Duarte highlighted the concern over risks such as fake information, job losses en masse, but also pointed out that AI would bring opportunities such as new types of employment, better management of society and health care benefits “enabling us to live better” have all been front and centre in recent weeks as AI has passed from the world of technology geeks to the mainstream. But is AI being hyped up or is it a game changer?
AI – not a match for humans yet
Professor Arlindo Olivera, fellow of Lisbon Higher Technical Institute – IST and President of INESC and an expert on AI said there was some hype but pointed out that it was a very important technology and stressed that AI was so much more than ChatGPT.
There are two important applications — analytics and automation — both with the capacity to replace people in certain areas such as translation and language.
“Intelligence is a fairly useful attribute, which allowed us to develop from hunting and gathering on the African Savannah to civilisation. In terms of knowledge discovery and value and product creation it seems it will be fairly useful over time,” he said referring to it as part of the 4th Industrial Revolution and which would save “so much time and bring so much value in revolutionising quality of life”.
Revolutionising health data
Pedro Marques (EPE) saw advantages in terms of health, from early disease detection, predicting pandemics, and medical imaging, although it would not resolve all the evils of health. But he said it could have an important role in Portugal in terms of data sifting and processing.
“Unfortunately here in Portugal our information system in the national health service needs a lot of work at a base level if we are to take advantage of all that AI can offer,” he said.
AI – a game-changer for a few
Daniela Braga (Defined.ai) said it would be a game changer but not right now. “I wouldn’t say ChatGBT is a game-changer, its simply all over the news, although it has gained more traction among users over the past two months than any other form of social network”.
“This is the democratisation of the knowledge and is the manifest existence of AI. This sudden spread (among a wider public) is new, in the hands of any citizen which it wasn’t before, and in a much more tangible language”, said the Defined.ai founder.
“It is a game changer for 5.0 Industry, and ChatGPT has given us a leap to artificial general intelligence (AGI) with computers on the same level as human intelligence, but it is a game changer for a few”, she added.
The world, the entrepreneur said, was “highly divided”, the venture capital investment was “still very slow and risk adverse since the pandemic”. Even (investment in) Web 3 and the Metaverse which had been going well had gone down with AI rising to the top with all this hype. “Investments in AI are hyped up being repeated with the same kind of investment in the same type of products to the order of US$100 million in pre-revenue companies. (Meaning a company that has not made any money yet from its productS or services)
It would be a game changer for those countries like the US and China which can deploy the technology and capital, but the rest of the world won’t be able to keep pace with investments of US$93Bn in the US and US$50Bn in China in 2022, while the rest of the world’s contribution is sluggish and almost non-existent by comparison.
Daniela Braga continued that the tech companies were leading the way in investment with the banking sector right at the bottom. Healthcare was leading more and showing more promise, particularly in disease and pandemic diagnostics, while investment was very considerable in the area of productivity which Defined.ai is so involved in.
“We can do 10 times more processing and content management, as well as responding to customers than we used to be able to do. However, there is a danger of information control, disinformation, and trust and safety and fraud. The other countries and sectors need to catch up with the US and China because talk isn’t enough. We need a national strategy and capital deployment plan,” she suggested.
Responsible and governable use
Madalena Talone (CGD) said the technology had existed for some years, not just in Large Language Modules* (*A large language model (LLM) is a type of artificial intelligence (AI) algorithm that uses deep learning techniques and massively large data sets to understand, summarise, generate and predict new content), but also Computer Vision* (*Computer vision is a field of artificial intelligence (AI) that enables computers and systems to derive meaningful information from digital images), for back office automisation and customer support – all various areas where AI has been used, but has become more visible recently with a growing number of use applications that is “within reach of all of us, cutting across all industries, and in which we have seen it quickly become tangible”.
“Faced with a technology that has been able to change a lot of things in a very structural and transversal way, we need to think seriously about the matter to ensure that we have the right mechanisms to be able to use these technologies responsibly and in a governable manner, but it will change a lot of things in our daily lives.”
Applications for banking
Manuel Maria Correia (DXC) referred to a study by McKinsey which pointed to a report that has forecast that the economic benefits of Generative Artificial Intelligence could bring economic benefits worth US$4.4Trn with three sectors bringing in US$1.2Trn, with one of these sectors being the banking sector.
Madalena Talone said it was only natural that the banking sector would be one of these sectors developing GAI since the sector was one by its very nature which “creates and processes a lot of information, enabling the banks to use the information in a more exponential way”.
She highlighted the increased level of care regarding confidentiality, information security, and a great responsibility in how this information was used.
“At Caixa want to do this in a manner that is very secure, responsible and in partnerships. We have created an Analytical Intelligence Centre to ensure we have the necessary skills to study the topic of AI in its various aspects like Modulation, Analytics, Task Automation, both for front and back office, Large Language Models and Computer Vision, all of which we have made a large investment on a fairly relevant team, and we think it is a topic with a lot of potential to be a game-changer in what we do while adapting this to the digital reality which is fairly new.”
Low Banking investment
Daniela Braga continued to insist that investment in AI in the banking sector, compared to other sectors, was inferior.
“Obviously each company and sector is making its own way, but in relation to the total investment per sector and country its peanuts because banking is a highly regulated sector, and digital transformation is not AI and AI is not Generative AI. To say we’re involved in digital transition may simply mean that in the banking sector that paperwork is now digital documents which very often it is,” she countered.
Ethics and regulation
Then there was the question of ethical AI and the checks and balances to ensure that these powerful tools are not misused.
“The question of ethics and regulation are different. Ethics is a concept of what is right and what is wrong – a moral concept – while regulation enforces this ethical code”, said Daniela Braga who added that AI was no different to any other sector or area that needed to be regulated.
“We have been in the forefront of this ethnical component because everything that is AI begins with data and we are the largest data market place in the world (the US) where data is ‘ethically sourced’ and we work with lots of partners who follow this code of ethics”, said the AI entrepreneur.
“Applied statistics” are not creative
Professor Arlindo Olivera discussed AI risk mitigation concerning data, but pointed out that all technologies carried risks, but saw four main areas of risk with AI: Information and false information, privacy, economic disruption, and the risks inherent from AI – which by the way was “applied statistics” — (“AI ending the human race sounds more dramatic than Applied Statistics will finish us”).
“This doesn’t mean however that this risk is complete nonsense, but the AI systems that we have at present are a lot less intelligent then we think, and not close to the intelligence of a human being”, he added.
“We are a long way off from AI systems which have creativity, they don’t yet know that they know, and can’t yet reflect on their knowledge and explain the rationale behind things, although this is being worked on,” he continued.
“I think the Language Modules are just the tip of the iceberg because when we have the language to be able to use and interact with accounting data bases, personal data bases, search engines, and engines for mathematical and physics symbols, or health etc, it will have a potential that is much greater than other languages. For example we’ll be able to consult with our digital accountant which will be able to tell us what are our financial business needs. I think that ChatGBT is just the tip of the iceberg in terms of the data these systems will have,” said the academic.
Applications in healthcare
Pedro Duarte (Microsoft) talked about the challenges Portugal faced in this area, particularly the public sector and the health system, pointing out that AI had appeared on the general scene at the same time as Portugal’s Recovery and Resilience Plan.
“One project we have is in the health area in Portugal, both in the private and public sectors, to have a single clinical register so if the user is referred to a private hospital the professionals can see what the patient did in the public hospital and vice-versa to end this dichotomy that exists in both sectors, replacing it with an integrated health care system.”
But this could only happen if they have all the data. “The good news is that the private sector was totally aligned with the public sector and I think that the advantages of AI here in this sector is fairly obvious.”
Portugal – carving a space in AI globally
But how could Portugal position itself against other countries like the US and China? Daniela Braga said she had spoken with the Portuguese government at the “highest level” and given its political neutrality and multiculturalism it meant that Portugal could have a place at the discussion tables more easily than when the US, Germany or the UK is leading the debate, because Portugal is seen in a different light economically and politically.
“We have this advantage that we get on with everyone, we have António Guterres at the UN, we have this set of security and geographical conditions, and we’re not seen as warmongers, which are advantages. I think Portugal has a lot to contribute, and we need to invest in creating an AI hub in Portugal and become specialised in this area rather like Switzerland is in the financial sector”, said the Defined.ai founder.
Manuel Maria Correia, Managing Director of DXC summing up the AI debate and if I was hype or a game-changer, concluded from all the comments that it was both – that it was a technology that carried a lot of risks with it, but that many of them could be mitigated, that there were opportunities in the health sector, for example, and while Portugal was too small to be a big player in AI, it does have a set of unique conditions to be an important player, and that these opportunities will be leveraged by economic sectors in Portugal, namely in the banking sector, but that this is a transversal topic and naturally there are concerns over how it will be policed.
Text and photos: Chris Graeme