AI in 2023 and beyond. A Friend or a Foe?

Sometimes it can feel like we are already living in a science fiction movie. A few years ago, it would have been unimaginable that you’d tell your fridge to play you a song or that your watch would remind you to take a walk. But this is only the start. The advancement of deep learning and natural language acquisition will speed up AI developments. The development of autonomous vehicles and weapons is already underway, and VR games and deep fake videos make it increasingly difficult to tell fact from fiction.

Many of us have hailed Artificial Intelligence (AI) as a technological breakthrough that could take almost all industries to unthinkable levels of efficiency. At its core, AI tools are designed to automate menial tasks, streamline operations, and improve overall productivity. But at what cost? And, no, this is not a question of how much you might be willing to invest in an AI app that can help your company; this is about the underlying threat these systems pose to the economy.

Back in the 18th and 19th centuries, when the First Industrial Revolution reshaped the entire economic landscape, people who could no longer farm because factories took over their land had to retrain themselves, learning how to fabricate or repair the same machinery that made them redundant. Before, there was little demand for these abilities, but jobs that didn’t exist in the past became a new alternative.

The Second Industrial Revolution saw the same scenario unfold. There may have been temporary job losses because of the switch from steam to electricity. Still, if those mechanics had become electricians, they would have kept their positions within their companies. Once more, there were brand-new skills that people could learn, and with the advent of mass production lines, there were more job opportunities than ever before.

So if history is any indication, we should be okay, right? While many economists point to how these past significant transformations in the labor markets did not lead to major social upheaval, I wouldn’t be so sure. According to these specialists, when technology eliminates jobs, people find alternative income sources, which is true. What they don’t specify is how long this transition takes and its damaging effects in the short term.

Can AI weaponize your data?

Deep learning has been one of the most significant AI developments in the last decade. Thanks to deep learning, computers can classify data, identify patterns, and make predictions. For example, Facebook uses deep learning technology to create personalized recommendations and maximize the time you spend on its network. The platform analyzes every click and compares your information to the data from millions of other users in their system to accurately predict what will keep you interested. The platform analyzes every click and compares your information to the data from millions of other users to accurately predict what will keep you interested.

Deep learning will only grow more prevalent and influential in the years to come, which is great because it can have enormous benefits in everyday life. AI can analyze vast amounts of data and draw connections that escape the human mind. Still, AI cannot use common sense, abstract concepts, or personal experience, so it lacks the complexity and nuance of human thought. To that point, in the podcast Growth Minds, Taiwanese author Kai-Fu Lee of the excellent book ‘AI 2041,’ explained that AI could not replicate human creativity yet, either.

In that same book, Lee also warns us deep fakes will get so good that it won’t be possible to tell them apart from the real thing. It took some time for developers to figure out how to get computers to process and comprehend images to create the technology for deep fakes. Therefore, they looked to the human brain for inspiration, which has a visual cortex that collects data about an image before sending it to the neocortex, which processes that data and then gives the image a more nuanced meaning. Then, they developed an AI that could accomplish this and succeeded. 

In short: Yes, your data can be weaponized. And no, it’s not only a matter of where you clicked or what you liked online—AI could even use your likeness against you if it falls into the wrong hands. 

What do the coming ten years look like for all of us?

The immediate impact of AI on employment is both positive and negative. While it can streamline operations, increase efficiency, and even generate demand for new jobs, it will also translate into widespread job losses and, at a minimum, significant wage reductions for those who remain employed. Chances are that the overall ramifications of AI over the next decade will be so complex that policymakers and organizations might need to create a legal framework to mitigate the damage to the economy.

However, waiting for any organization to take the reins is a fool’s errand. AI will change how the entire workforce operates. Individuals and businesses must be proactive and prepare themselves for these changes by acquiring additional skill sets, embracing these advanced technologies, and adapting to the new reality.

What are the major threats?

At first, one of the most significant concerns with AI was that it has the potential to automate low-skilled tasks that are repetitive and require minimal human interaction. However, as AI systems have become more sophisticated, it’s becoming clear that these apps can perform other, more “creative” jobs almost as effectively as humans, but in a matter of seconds.

In just a few years, will we need as many designers or copywriters in advertising as we do today? It’s hard to tell. At this point, it is not uncommon to hear of AI algorithms in the financial sector that have automated routine tasks, such as data entry, invoicing, and payment processing. This resulted in countless layoffs in this industry, but it wasn’t alarming yet. Now that AI could displace workers in higher-skilled roles, and at a larger scale than ever, people across many verticals are losing their sleep, wondering if their jobs might not even exist in a few years.

Over the last two decades, most AI has focused on analyzing already-existing information, such as finding data anomalies, spotting fraud, or recommending products you might like. However, generative AI is an entirely different ball game because It helps you produce original content.

With well-known text-to-image AI tools like Stable Diffusion or DALLE, you simply need to enter a prompt to produce images that are difficult to distinguish from works of art and photographs created by humans. Once more, these may even be in a specific artist’s style, which has already led to a class action lawsuit.

Enter ChatGPT

When it launched in November 2022, ChatGPT sent shockwaves through the tech industry as users were astounded by how well it could comprehend and produce conversational responses. This bot, developed by San Francisco-based research laboratory OpenAI, is capable of a seemingly limitless number of tasks, including writing a college essay, answering questions, creating and debugging code, and creating stories in a particular style. 

Powered by machine learning, this technology has revolutionized the field of language processing, allowing it to generate human-like responses to a wide range of queries and prompts in the blink of an eye. ChatGPT is a conversational chatbot that, in its own words, “uses deep learning algorithms and a massive amount of training data to generate humanlike responses to text-based prompts.”

ChatGPT has been a revelation for many people because it demonstrates a level of artificial intelligence that previously seemed unattainable. After all, we still get frustrated with Siri and Alexa over the most straightforward requests. So it is fair to say that the embryonic version of this AI has been a success, so much so that Microsoft reportedly invested $11 billion in OpenAI in January 2023, and it is considering adding ChatGPT to Bing in an effort to dethrone Google.

Search engines are about to change

Whether you know it or not, artificial intelligence is already prevalent in many products you already use. That’s not the novelty. The main difference is that ChatGPT has brought AI front and center by enabling users to chat directly with it. Moreover, ChatGPT has proven that, since it can succinctly explain complex concepts, respond to questions, and combine information from different online sources, it could be better at performing many tasks that people use search engines for every day.

As stated before, this potential hasn’t gone unnoticed by Microsoft. If the heavy investment wasn’t indication enough, Microsoft’s CEO, Satya Nadella, has said that their company sees ChatGPT as an opportunity to take on Google. On the other hand, Google has announced Bard, its own chatbot that will work as a “companion” to its search engine. Additionally, Google has invested $300 million in Anthropic, a company started by former OpenAI employees and home to the chatbot Claude. And Baidu, the “Google of China,” also announced Ernie, a chatbot, that would be released in March.

But are these chatbots as reliable as Google or Bing search results?

No, they are not. ChatGPT frequently makes mistakes. Chatbots can present their responses as gospel truth, unlike search engines that point users to other pages and make no claims regarding their veracity. Chatbots are sometimes prone to bias, prejudice, and false information as they learn from various online sources. This is where tech companies must decide where to draw the line as they determine which topics are too delicate. All of this will lead to concerns about truth’s nature, objectivity, and censorship.

How dows ChatGPT work?

Deep learning uses neural networks, essentially matrix equations, to gather and learn the relationship in that data to produce what is referred to as a model.

Because of its unparalleled language generation capabilities, ChatGPT can be fine-tuned to the specifics of the task at hand, making it a highly customizable tool that lends itself well to various industries, including customer service, marketing, and even journalism. And, of course, that combination of accuracy and speed is what makes this bot so appealing and, frankly, terrifying. Since it can generate accurate real-time responses indistinguishable from a human’s, no one can deny its cost-effectiveness.

Here you have the ideal customer support representative, consistently providing quick answers, 24/7, without breaks. Sounds too good to be true? It might be

Is ChatGPT a tool with unlimited potential?

Technology does have its limitations. The data only extends to 2021. Responses can be monotonous, lack warmth, based on false information, and occasionally devolve into a word salad. Due to its machine learning model limitations, ChatGPT could perpetuate biases in the data it consumes to learn. This means it may provide discriminatory or offensive responses that could aggravate your clients.

Despise its promising early results, ChatGPTalso struggles to understand the nuances in conversation, mainly if the prompt is ambiguous or the user presents an open-ended question. So far, this has resulted in strange interactions with some users that range from off-topic to irrelevant to creepy. However, Chat GPT keeps track of questions and answers throughout a conversation, allowing you to refine them.

Will there be anyone left?

AI can improve the accuracy and speed of decision-making in critical areas, such as finance and healthcare. You could surely offer more personalized products or services with the right AI tools and create entirely new business models. That’s all well and good. But if you automate every process that you can, and every other company does the same, then who will be left? In other words, who will pay for what you offer if no one has a job except for a selected few? Because there’s something that AI will never be able to do — become your customer.