Hola MKAI champions!
In this week’s spotlight, we emphasise how learning and gaining knowledge can bring us closer to understanding ethics and meaning in AI.
Latest reflections on AI ethics
Before bringing AI discussion to the table, let’s dive into the deep end of theory, then we can go back into the real world and find practical uses for it. We have prepared quite a few treats for you this week.
Although there have many attempts to measure ethical behaviours and decision making, there remain many questions about what we are precisely trying to measure or quantify. An example:
We are seeking “fairness” in AI, but can we discover any absolute truths behind fairness?
Ethics and Meaning of AI
Inspired by the article about the Ethics of AI: Benefits and risks of artificial intelligence, we are starting from the hypothesis that everything is relative. There is this duality such as good and evil, benefits and risks, but before that, let’s start with fairness behind it. The truth behind being fair is empowered by the subjective inner feeling we possess.
The subjective inner feeling is our unique gift that AI still doesn’t have.
And whether what we believe will bring positive or negative aspects depends on the perspective from which we look at it. If we think we are on the right path to establishing ethical artificial intelligence, it will bring good. If we do not believe in it and look from a negative perspective, we will manifest such a reality.
At the moment, we don’t have a single piece of evidence, except for the inner feeling, because the truth is a subjective sense of fact, not absolute truth.
Hidden layers and their historical relevance
Even though conversations from the past can bring so much knowledge to our present selves, there’s still a question or concern of the validity of historical data.
What is the absolute truth?
Yes, we can remember and interpret what has been happening during the last week or maybe even month, but should we examine the validity of historical data?
That’s why human history should not be objective because we wrote it, humans. And machine learning algorithms are not conscious of these facts. They’re just doing great work when it comes to recognising patterns in historical data to form predictions.
Then, rather than encouraging fairness and ethics, they can instead encode bias. The results can be both encouraging and disappointing.
For example, if we have so many advanced algorithms that can recognize fake news and biased interpretations, how can we be sure that the old transcripts and hearsays are accurate? It is not about conspiracy theories but whether we will be able to give algorithms accurate data. Therefore, it is evident that AI needs to become an integral part of the new media world.
Expressing the contradiction of good and bad effects of AI, Steven Mills writes in the preface to The State of AI Ethics that artificial intelligence has a dual nature:
“AI can amplify the spread of fake news, but it can also help humans identify and filter it. Algorithms can perpetuate systemic societal biases, but they can also reveal unfair decision processes. Training complex models can have a significant carbon footprint, but AI can optimize energy production and data center operations. The challenge for us as a community is to minimize the unintended consequences and shape a future in which AI will be used to right wrongs, deliver positive solutions, and increase optimism and hope.”
Algorithms have higher ethical standards than humans.
Regarding dualism, there is still one more concern. Even if we provide algorithms the clean data, we cannot expect AI to be moral and honest if we don’t respect our ethical framework. Of course, it’s the beauty of humanity that we all have different values across diverse cultures. Still, the question is related to – How can we expect AI to be enlightened, moral, and honest if we don’t act ethically?
Therefore, it’s essential to put together groups of people from different backgrounds and include academic and NGO perspectives so that conversation about ethical AI is multisectoral. In fact, this is the MKAI mission statement. If you have your view on this topic, please join our Telegram for more focused discussion topics like this one.
The entire universe might be a neural network.
As we mentioned above, truth is a subjective sense of fact, not absolute truth. The same goes with this crazy idea that might be crazy enough even to be true. And it’s about the fact that the entire universe might be a neural network.
Let’s look at this from a new angle. It’s fascinating how information shared between different regions of one’s brain is similar to how information shared among people from all walks of life is saturated in collective intelligence. We are aware that we are connected, but – what happens when intelligence accelerates beyond evolutionary sense?
This is thanks to synapse plasticity, which is the basic building block of the brain’s computational power. These synapses enable the brain to work in a highly parallel, fault-tolerant, and energy-efficient manner…mimicking critical functions of a biological synapse.
But it seems that we are coming to that part to go beyond the biological framework. The article related to a Brain-like device that mimics human learning in major computing breakthroughs raises the question of how we use advanced computational capabilities.
Quantum computing leads to building better AI systems
Conventional computers store data and process data using separate systems, meaning data-intensive tasks consume a lot of energy. That’s where AI can help. Or better to say quantum AI.
Step closer to Quantum AI and quantum algorithms
The code of native content computers is a bit like assembly, not easy to write. And for humans, it feels like the 70s are back, and they require a whole new mindset and perspective needed for that. That’s where AI can help computers write that code, and that’s how we can use native computers to improve AI models.
Still in the mood for more knowledge to digest?
How close are we to having powerful, fully programmable quantum computers?
For instance, one of the things that we can do with today’s quantum computers, we can perform some random number generation, something that an average supercomputer cannot do quickly. In the short term future, applications will not be used widely, but quantum technology 2.0 promises a completely new quantum algorithm when it comes to coding. New types of hardware will be produced for quantum computing. Instead of using a modified classical algorithm, you’ll be able to use quantum algorithms only.
The introduction about quantum computing and entanglements was essential because they are related to AI.
Regulation is necessary; establishing ethical frameworks is urgent because we are direct witnesses of the advancement of AI systems. And as far as the entanglement is concerned – we are facing enormous challenges.
Right now, some governments take artificial intelligence risks seriously, as it is a matter of national security and defence. However, issues connected to AI aren’t related to defense and security only; it is actively affecting all aspects of society, starting from education, criminal justice, medicine, the economy, etc.
So, what does this mean for quantum computing and intelligent systems? What are we going to do with it that we cannot do with current computational resources? Are these issues worth the effort? Surely intelligent systems will deliver a new way of performing parallel analysis.
Hence, are we doing anything new with intelligence – engineering better genomic therapies, etc.? Or there is some root and history is being recreated, since we are in a geometrical progressing graph that is not too accurate.
The brain wants meaning before the detail.
“For humans, honestly, the outer journey is the easy bit. We can go to mars. We can colonise the galaxy, we can terraform planets and we can monitor everything on them and on earth. We can get everything working exactly as we want it. We can live in harmony from an external perspective. Still, the inner game, the inner journey that’s our fundamental challenge – to get past all the superficial stuff and find that higher level of consciousness, that higher level of love, that higher level of peace. That’s ultimately only ever going to be what our human story is about.“
– Richard Foster-Fletcher; Source: Singularity Watch S01 E13 | Virtual Reality & Artificial Intelligence will change your life
What’s your language – expressing experience and love
We are not alone. If you thought that only humans could communicate with each other, you might make up your mind again. Scientists have found that above-ground stresses in plants have resulted in complex communication systems (electrical, mechanical, and chemical) that influence below-ground communication between neighboring plants.
Graphical illustration of above-ground interactions between neighboring plants by light touch and their effect on below-ground communication. Credit: Elhakeem et al (2018)
This was explained further in Groundbreaking Research on Plant-to-Plant Communication by Christian Nansen, where he emphasized that even plants don’t have neurons or brains, respected scientists approach them like they have one. The last decade brought not only significant progress in the field of plant-plant communication. It also brought crucial demonstrations that plants can respond to volatile aerial compounds, root secretions, and sound. You can read more about it in the latest research paper, published by Alexander Volkov, Underground electrotonic signal transmission between plants.
Decoding animals – how much will we be able to understand them?
Speaking of plants, when was the last time your cat knocked down something from the table? Or dug the ground out of the pot where your plant was resting and minding its own business? Well, you might be able to see why and discuss some questions with your pets shortly.
Kudos to researchers that are using machine-learning algorithms to decode this precious communication that is occurring among animals in their language. This can connect humanity with intelligent species such as dolphins, elephants, and our fellow great apes. Note the emphasizing of we understand them, not the other way. However, there is a reasonable doubt that genuine translation is possible between species that don’t share fundamental perceptual and cognitive processes.
While we may use ML to understand animal language, the question is whether our understanding is based solely on human interpretation of animal language. They may understand us, while we do not understand them. More important – do we even understand ourselves? What are our love languages?
Upcoming MKAI events – Focus on AI-driven data monetization
How do we monetize our data, and how do we monetize our content?
First and foremost, there is a question arising on will we be able to process large amounts of data in the next 10 years of so. Even if we could do that, there is a quality assurance concern, related to how well we’ll process that data. The answer may lie behind focusing on transparency and data quality while applying advanced analytics and AI to unlock its massive potential.
What is required for companies and organizations to survive and then thrive is extracting real value from data and using the specific information for making further business decisions. It’s time to expand our views and evolve from current data strategies.
At the May MKAI Inclusive AI Forum, we’re going to discuss the exact topic of ‘Rethinking Data Monetization.’
The Forum is inclusive, digital, and on Thursday 27th May 2021. Get in touch! We’d love to show you the impact of rethinking data monetization!
Not yet a part of our community? Come and join us!
We hope you enjoyed this week’s selection! If any of the topics left a particular impression or resonated with you – feel free to join our focused discussion group on Telegram.
See you next week,
Alex
