-
Asked by anon-379478 on 8 Jan 2024.
-
Alexander Coles answered on 8 Jan 2024:
In all science and research fields we build on previous work done by those before us. On the common £2 coin you will see words “STANDING ON THE SHOULDERS OF GIANTS” engraved on to its side. This means that we are able to continue to do new advanced research only thanks to those who have done so before.
-
-
Andrew Maynard answered on 8 Jan 2024:
AI is becoming increasingly advanced through thousands of people working together on challenges and learning from each other — as well as building on what did and didn’t work in the past. AI has been around for decades, but breakthroughs over the past few years have enabled researchers around the world to make amazing advances in what is possible.
-
Muhammad Malik answered on 9 Jan 2024:
It was not an overnight success. The success is a virtue of past efforts and achievements, especially by computer technology developers. So, yes the advancement in AI is a result of collaborative efforts by a large team of scientists and engineers working together to make AI smarter and better, leading to remarkable progress/tech which the world is benefiting from today.
-
Mohan Sridharan answered on 9 Jan 2024:
The current state of AI is the result of decades of effort by researchers around the world. At the same time, I would like to point out that the capabilities of AI are not as advanced as being described in popular media 🙂 There are still many open problems offering scope for a lot more research.
-
Gareth Hartwell answered on 9 Jan 2024:
Like most scientific developments it’s a combination of a few people coming up with some really clever ideas (such as realising that in principle you could ‘breed’ computer programs a bit like animals – which happened a very long time ago and more recently the pattern recognition ideas underlying Large Language Models such as chatgpt) and lots of people working very hard on improving and evolving the models. The improved power of computers over this time has been a very important factor too. University groups have now been working on this for over 30 years.
But maybe it isn’t yet as advanced as you might think. Don’t believe everything the people selling AI platforms tell you and try out for yourself how good they are and find out what they are good at and what they aren’t.
-
Rosemary J Thomas answered on 9 Jan 2024:
One of the most important advances in AI is called machine learning, which means making machines that can learn from data and improve themselves without being told exactly what to do. Machine learning can use different methods, such as deep learning, which means making machines that can learn from many layers of data and find patterns that humans might not see. For example, deep learning can help machines to create realistic images, understand natural language, or play complex games.
AI can be used for many things, such as helping doctors diagnose diseases, helping students learn better, helping farmers grow more crops, helping artists create new music, or helping scientists discover new things. AI can also help us to have fun, such as chatting with us, telling us jokes, or making us laugh. AI can be very helpful, but it can also be very powerful, so we need to be careful and responsible when we use it.
-
Carl Peter Robinson answered on 9 Jan 2024:
Research into AI has been around for many decades and that research gets pushed forward with each new generation of scientists that come into the field. It should be noted that the successes and results achieved in roughly the last 10 to 15 years are mainly down to the relatively recent advancement in computer hardware. Scientists realised they could use graphical processing units (GPUs) to run their AI algorithms and models much quicker than had previously been possible. This sped up scientists’ ability to try out new experiments and test their theories. Now there are companies like NVIDIA who build dedicated hardware to run AI models on. This is one of the main reasons why AI has advanced so rapidly recently: we now have the hardware and resources to train much larger models and at a quicker pace. You often find scientific progress and breakthroughs occur when an improvement happens in humans’ capability to take measurements.
-
Sheridan Williams answered on 9 Jan 2024:
It uses ‘machine-learning’ a technique based around neural networks. It collects information by trawling the internet and other locations for information. Some publishers are annoyed that information it has collected from them is used without credit or payment.
I typed your question into ChatGPT and here’s what it said…
The advancement of artificial intelligence (AI) is a result of several factors and ongoing research and development. Here are some key contributors to the advancement of AI:Data Availability: The availability of vast amounts of data has played a crucial role. Machine learning algorithms, a subset of AI, often require large datasets for training to improve their performance.
Computing Power: Advances in hardware, particularly the development of powerful GPUs (Graphics Processing Units) and specialized chips for AI tasks, have significantly increased the computational capabilities required for training complex models.
Algorithms: Continuous refinement and development of algorithms, such as neural networks and deep learning techniques, have greatly enhanced the ability of AI systems to understand and process information.
Research and Collaboration: The AI community is actively involved in ongoing research, and collaboration among researchers, scientists, and engineers has accelerated progress. Open-source contributions and the sharing of knowledge have become common in the field.
Interdisciplinary Approach: AI development often involves collaboration between experts from various fields, such as mathematics, computer science, neuroscience, and cognitive science. This interdisciplinary approach helps in creating more holistic and effective AI models.
Feedback Loops: Systems that can learn from their own outputs and improve over time, known as reinforcement learning, have contributed to the development of more autonomous and adaptable AI systems.
Increased Funding: Growing interest in AI from both public and private sectors has led to increased funding, enabling researchers and organizations to invest in more ambitious projects and experiments.
Real-World Applications: The deployment of AI in real-world applications, from healthcare to finance and beyond, has provided valuable feedback and insights, driving further improvements.
The combination of these factors has led to the rapid advancement of AI technologies in recent years. Ongoing research and collaboration across disciplines continue to push the boundaries of what AI can achieve.
-
Fraser Smith answered on 9 Jan 2024:
There are lots of people working on making AI systems better but much of the recent success comes from having access to lots of data to train on (millions of images to be able to recognize what’s in them – e.g. a specific type of dog) or much of the text on the internet (for chat GPT). Together with extremely powerful hardware this drives the recent successes.
-
Luke Humphrey answered on 9 Jan 2024:
This is a really interesting question! I’m not sure it’s possible to say there’s one reason AI has got so advanced as it has. The origins of machine learning are in the 1950s, but the underlying maths is much much older.
One major factor has been the rapid development of computers over the last few decades: as computers have been more or less doubling in computational power every year since the 1970s and have only just started to show signs of levelling off now in the early 2020s.
To put it in perspective, the guidance computer used to land on the moon in 1969 had 4kB of RAM and weighed around 30kg. In 2024, a typical phone has upwards of 8GB of RAM (about 2000x more) and weights about 250g (about 12000x less).
AI requires a lot of computational power to work effectively, so one reason that it has got so advanced so quickly is linked to the more general rapid advances in computer power.
And now, as AI becomes more popular and the market for AI computing grows, we are seeing hardware change to match. For example, GPUs (graphics processing units) have typically been driven by the gaming market but now we’re starting to see vendors build GPUs that are not optimised for rendering graphics but for training AI models.
This sort of thing is common with all new technologies and is called the “S curve”. If you draw a graph of the rate of progress over time: it’s flat at first, then gets steeper and steeper upwards as people adopt and invest in the technology, then it flattens off again once the new technology becomes fully established – so it kind of looks like an S. For AI, we’re still at the bottom of that S curve as it grows steeper and steeper and there’s really rapid advances as more people get involved.
-
-
Mackenzie Jorgensen answered on 9 Jan 2024:
It learns from data that scientists train it on or from trying out different actions and seeing what rewards it receives. AI is never perfect and it will make mistakes no matter how advanced it gets–it’s just a tool at the end of the day.
-
Kevin Tsang answered on 10 Jan 2024: last edited 10 Jan 2024 12:12 pm
I think there are three aspects to this.
1) The field of AI overall has recently advanced very fast due to the development of computing power and GPUs, where machine learning has become feasible. This has allowed machine learning and specifically deep learning algorithms to be more powerful. Remember that machine learning is one way of creating AIs.
2) To train a specific machine learning algorithm to do a task in the real world, we feed it lots of relevant data about the task so that it can be trained. This large amount of data has only been made available in the past 10 years or so. By learning the many patterns in this data, the AI is able to give an advanced representation of the real world.
3) There is more interest in AI, leading to more research funding and people working on developing AI technology.
-
Demetris Soukeras answered on 19 Jan 2024:
Many scientists and engineers working hard improving on each other over the last century.
What has also helped a great deal are the improvements made to computers, as computers get more powerful we can make more complicated AI.
Comments