Echo Buzz: The Rise of AI

The Spark

Most of us were probably introduced to the concept of AI through films such as 2001: A Space Odyssey (1968), The Terminator (1984), The Matrix (1999), and Wall-E (2008). These films typically pushed the narratives that 1) AI is a thing of the future, 2) AI will turn against humans, and 3) AI will replace mankind. Thus, although these films can be credited for introducing the concept of AI to the masses, they are guilty of reinforcing negative stereotypes about AI and sowing the seeds of distrust among the people.

AI isn’t exactly a novel concept as its roots can be traced back to ancient civilisations such as the Greeks and Egyptians who believed that the human-esque statues they sculpted possessed a mind capable of reason and emotion. Meanwhile, renowned 17th-century philosophers such as Descartes, Leibniz, and Hobbes broke new ground with their visions of a universal language of reasoning that would make rational thought as black and white as mathematics. 

However, it wasn’t until the 1950s that AI became more than just a far-fetched fantasy, but a plausible reality. This was largely due to revolutionary scientist Alan Turing (yes, the same Alan Turing played by Benedict Cumberbatch in 2014’s The Imitation Game), who posed the question: Why can’t machines be trained to solve problems and make decisions the same way as humans? Yet, Turing couldn’t do much with this revelation as computers were still in their primitive form at the time (and out of his price range). 

The Bloom (and The Gloom)

It was only a few years later that AI reentered scientific discourse with the historic Dartmouth Workshop of 1956, which introduced the first AI program and set the pendulum into motion for the field. Things were off to a promising start with the invention of the General Problem Solver which was meant to find the solution to any problem in the universe. Clearly, it didn’t live up to its name. This was largely due to the fact that the machine was designed to navigate an endless maze of infinite possibilities when approaching a problem; pausing at each intersection to make a decision and backtracking every time it reached a dead end, making it impossible to arrive at the right solution. Researchers also attempted to dip their toes into natural language programming with the invention of ELIZA – the ancestor of today’s AI chatbots. And, although ELIZA was able to hold a conversation with a human (and even act as a therapist), its responses were often meaningless or stilted. 

“Machines will be capable, within twenty years, of doing any work a man can do,” – H.A. Simon

Despite these half-baked attempts, researchers were still quite optimistic about their odds and had fairly high expectations for what they could achieve. Eventually, their efforts were vindicated when government agencies started pouring their money into AI research institutions. This spurred their research, leading to the invention of Shakey the Robot which could break down complex instructions into individual commands, and WABOT-1, the world’s first full-scale anthropomorphic robot. But, they hit a brick wall just as they were gaining momentum due to limited computer power- computers couldn’t store sufficient data or process information fast enough – and a lack of funding – investors pulled out once they realised researchers couldn’t meet their ambitious goals. 

Then, came the RenAIssance. Thanks to the expansion of the algorithmic toolkit and an injection of funds, scientists were able to develop programs that used the deep learning technique, where computers learnt by building on previous experiences as opposed to collecting information in a linear fashion. Another form of AI that gained popularity in the late 80s was the expert system, such as Japan’s Fifth Generation Computer Systems which quizzed experts from various fields to equip its database with answers to pretty much any problem. AI continued to work quietly in the background for most of the late 80s and early 90s, with researchers making slow but sure progress in its development. Although it was already shaping into a formidable force in industries such as financial services and medicine, its presence mostly went unnoticed by the general public, until now.

The Boom (and The Doom)

“A lot of cutting edge AI has filtered into general applications, often without being called AI because once something becomes useful enough and common enough it’s not labelled AI anymore,” – Nick Bostro

Tick, tick…BOOM! The spark has finally ignited, with AI going from a silent player behind the scenes to a bona fide star seemingly overnight. Yet, you might be surprised to find out that we’ve spent almost a decade indirectly interacting with AI. Whether it’s asking voice assistants like Siri or Google Assistant about the weather, unlocking your phone with facial recognition, driving somewhere with directions from Waze or Google Maps, scrolling through the For You page on TikTok, watching a recommended film on Netflix or Disney+, listening to a mix made by Spotify, playing against a bot in a Dota 2, or just searching a random fact on Google; you’ve had contact with AI! 

But, these forms of AI merely exist to make our day-to-day lives easier and are hard-coded for specific functions that developers perceive to be of the most use to consumers, struggling to perform more complex tasks. Enter OpenAI; OpenAI is partnered with and largely funded by Microsoft in an effort to integrate AI systems into their services to find different ways in which they can improve their already existing services. Their innovations have propelled them to the top of the industry, competing with the likes of Google, which before was perceived as unmatched in the field of AI. OpenAI was only established in 2015, which goes to show how impressive its meteoric rise has been.

Open AI models such as ChatGPT and DALL-E 2 have recently taken social media by storm as they are versatile and powerful while still being accessible and user-friendly. ChatGPT is more than just a chatbot; it’s a natural language processing tool that interacts with users in a conversational way to answer their questions and solve their problems. To get a clearer picture of the types of tasks performed by ChatGPT, we went straight to the source, and here’s what ChatGPT had to say:

1. People often ask me to provide information on a wide range of topics, from history and science to current events and pop culture.
2. Users often ask me to define words or concepts that they’re unfamiliar with.
3. People ask me to translate words or phrases from one language to another.
4. Users often ask me to help them with writing tasks, such as drafting emails, essays, poetry, stories or resumes.
5. People ask me to help them solve math problems or explain difficult concepts.
6.  Users often ask me for advice on a wide range of issues, such as relationships, careers, or health concerns.

Of course, we had to put ChatGPT to the test and see if it was as good as its word:

1. General Knowledge Questions

Q: How long did World War II last?

A: World War II lasted from 1939 to 1945. It began with the invasion of Poland by Germany on September 1, 1939, and ended with Germany’s surrender on May 7, 1945, followed by Japan’s surrender on August 15, 1945, after the dropping of atomic bombs on the Japanese cities of Hiroshima and Nagasaki.

Verdict: Gave us facts that weren’t relevant to the question and made us do our own math to figure out that it lasted for almost six years. So, we’re off to a pretty disappointing start.

2. Definition requests

Q: Can you define Artificial Intelligence (AI)?

A: Yes, of course! Artificial Intelligence, commonly abbreviated as AI, is the simulation of human intelligence in machines that are programmed to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and language translation. AI systems are designed to learn from the data they receive and improve their performance over time. Some examples of AI applications include virtual assistants, image and speech recognition software, self-driving cars, and predictive analytics tools.

Verdict: Not bad, though we’d be pretty surprised if it couldn’t answer this one. We love the human touch of ‘Yes, of course” and the response would be useful for anyone who’s new to AI. But, a quick Google search would have done the trick if we were being honest. 

3. Translation Requests

Q: Can you translate “May the force be with you” to Malay?

A; “May the force be with you” can be translated to “Semoga kekuatan bersama kamu” in Malay.

Verdict: Definitely better than anything we could come up with, but it doesn’t exactly roll off the tongue. However, Google Translate does the same thing, so we’re not entirely convinced of ChatGPT’s prowess just yet.

4. Writing Assistance

Q: Can you write a short poem on AI?

A:

AI, the future is here in sight,

Machines that learn, machines that might,

From Siri’s voice to self-driving cars,

It’s come so far, reaching for the stars.

It predicts, it helps, it makes us see,

A world of data, it sorts and sifts with glee,

But fears abound, that it may take,

Jobs and power, a world at stake.

We must use it well, to benefit all,

And ensure it never makes us fall,

For in the end, it’s we who create,

The future with AI, our fate.

Verdict: Well, colour us impressed; a poem, in less than a second! Looks like Echo’s writers may need to consider retirement at this point. Yet, we have to admit that though accomplished from a technical standpoint, it lacks the heart that only a person can put on paper.

5. Math Problems 

Q: Solve ∫ e^(2x) dx

A: Sure! To solve ∫ e^(2x) dx, we can use integration by substitution. Let’s let u = 2x so that du/dx = 2 and dx = du/2. Then we can write the integral as:

∫ e^(2x) dx = (1/2) ∫ e^u du

Now we can integrate e^u with respect to u, giving:

(1/2) ∫ e^u du = (1/2) e^u + C

∫ e^(2x) dx = (1/2) e^(2x) + C

So the solution to the integral is:

∫ e^(2x) dx = (1/2) e^(2x) + C

Verdict: Math teachers, meet your worst nightmare! This would be a lifesaver for kids as ChatGPT doesn’t just provide an answer but gives us a fully-worked solution. We’re starting to see why there’s so much hype now.

6. Advise Requests

Q: Can you provide advice to people who might lose their jobs because of you?

A: My advice is to focus on developing skills that are difficult to automate such as creative problem-solving, critical thinking, emotional intelligence, and communication. Stay up to date with emerging technologies and be adaptable to new roles. Remember that AI is a tool that can make our lives easier and by using it to our advantage, we can create new opportunities and build a better future.

Verdict: Easy for you to say, ChatGPT! Impending doom aside, it is pretty solid and ChatGPT does get more specific upon further prompting. 

Well, we can definitely conclude that ChatGPT definitely lives up to the hype as it helps us work faster and better. The news has been flooded with articles of people finding new and interesting ways of ChatGPT solving their problems. Software engineers use ChatGPT to write lines of code for them, and students use it to help them write their essays and assignments (much to the dismay of their teachers). Most impressively, ChatGPT passed job interviews at some of the biggest companies in the world and subject exams at various prestigious universities.

However, it’s still in the primal stage of evolution and has many limitations. Even OpenAI admitted that ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers and suffers from algorithmic bias as its answers tend to be skewed towards left-wing ideologies. Presently, the model is also incapable of handling heavy amounts of user traffic and has a tendency to crash. Some have even raised concerns that the model will increase the dependence of humans on AI and will not be able to come up with original content. For example, ChatGPT has been accused of acting as a scammer’s playground as it can be used to write convincing scam emails. These developments are especially concerning in the wake of Microsoft’s decision to lay off its ethics and society team in the AI division.

On another note, OpenAI’s DALL-E 2 has also been gaining traction on social media platforms. DALL-E 2 is an AI model that can generate realistic high-resolution images and art based on text descriptions provided by the user. DALL-E 2 has been praised for its ability to generate extremely detailed images and replicate various art styles that range from pop art to impressionism. 

Clearly, these images which were generated in less than 20 seconds are incredibly accurate to their descriptions. But, what could they possibly be used for? Although many people generate images just for laughs, they can be incredibly useful for artists seeking inspiration. But, it has caused an outrage in the art space, and rightfully so as DALL-E 2 raises the question: Since AI-generated art can be produced in a shorter period and at a lower price, do we still need artists? AI art generators like DALL-E 2 could reduce demand for actual artists, and rob them of their livelihood. Hence, it’s no surprise that artists have spoken out against AI art as training AI models use their original work, claiming that it appropriates their unique styles and infringes on their copyrights. 

The Question Mark

So what does the future of AI look like? Well, for now, we’re still a ways away from fully sentient beings capable of forming their own thoughts and beliefs. Nevertheless, it’s certainly worth keeping an eye on as it has the potential to change, well, everything we perceive to be possible. Technology is central to our lives in the 21st century; we live and breathe it, and it encompasses every aspect of our lives. Any significant change towards it is sure to have immense ramifications on how we live our lives moving forward.

Currently, the goal seems to be shifted towards just improving and refining conversational AIs like ChatGPT and working on integration with already existing services. Microsoft has recently integrated ChatGPT into its own browser Microsoft Edge, acting as a “co-pilot” to your web surfing as put by Microsoft. This poses an immense threat towards Google which for the longest time has dominated the internet browser space without contention. In response, they’ve tried to rush out their own version of ChatGPT called Bard, which they unveiled in a recent reveal event in order to compete with Microsoft’s recent announcement. Long story short, the reveal didn’t go very well due to the feature being forced out the door before it was ready and Google’s stock price crashed afterwards.

It’s been a while since Google has had to face any real competition within this space. This recent development will hopefully bolster and reinvigorate the industry to push harder towards greater technological developments and maybe in our lifetimes, we may all have a JARVIS of our own.

References:

1. The Rise Of Artificial Intelligence: Will Robots Actually Replace People?
2. The Timeline of Artificial Intelligence – From the 1940s
3. The History of Artificial Intelligence – Science in the News
4.  What is ChatGPT and why does it matter? Here’s everything you need to know 
5. OpenAI
6. How ChatGPT can revolutionize the way we solve problems
7. Microsoft unveils updated Bing and Edge with ChatGPT integration

Written by: Ethan and Priyanka

Edited by: Caitlin