3 Forecasts for Artificial Intelligence in 2024: An Analysis

AI won’t lead to our demise in 2024

Throughout 2023, we’ve witnessed several waves of fear, beginning with the April open letter I endorsed, urging a halt to “Giant AI Experiments.” Despite this, AI has continued to advance, albeit now under emerging regulatory scrutiny.

Another wave of concern emerged over OpenAI’s mysterious “Q*” technology. Curiously, OpenAI seems to be adding to the hysteria, with Steven Heidel of OpenAI tweeting, “Brace yourselves, AGI is coming” just recently. However, the actual news turned out to be about a “preparedness framework” rather than a breakthrough, leading me to believe it’s more about generating buzz than a real threat.

Here’s why I’m skeptical of the apocalyptic predictions:

• I don’t believe AGI is on our immediate horizon. Yann LeCun of Meta has even said current AI isn’t quite at the “dog level.”

• Even if we do develop advanced intelligence, it doesn’t necessarily mean it will have the desire or capability to dominate or destroy humanity, as machines lack emotions and desires.

• Focusing too much on AI as a harbinger of doom distracts from actual AI-related issues, such as intellectual property theft, the spread of fake media, and the potential displacement of jobs, particularly for freelancers.

The Second Wave of Generative AI is Approaching 

Indeed, 2023 marked the dawn of the first wave of Generative AI, with ChatGPT standing out as a notable milestone. The year was characterized by a competitive rush to produce the most advanced chatbot, led by OpenAI, with Google closely trailing with Bard among other innovations.

We also witnessed the emergence of the first wave of image creation tools, with Midjourney taking the lead, followed by entities like Stability A.I., DALL-E 2, and more. Even the code-assisting tools for developers were part of this generative wave.

However, this initial wave was more about showcasing potential through demonstrations rather than offering fully-fledged products. Demonstrations are meant to impress and prove concepts, while products are developed for practical use. I’ve explored this distinction in my post comparing Notion AI and ChatGPT.

As we move forward, the era of chatbot demonstrations is transitioning. While they’ll continue to evolve and improve, 2024 is poised to be the year of specialized AI-integrated products and services.

A notable subset of these upcoming offerings includes “enablers” like GPT-4 and Gemini, which other companies utilize to create end-user products.

The array of startups venturing into AI-powered solutions is extensive, offering services in various sectors like marketing (Smartly.io), customer service (MaestroQA), e-commerce (Lily AI), travel planning (“Everyday Agents”), and copywriting (Jasper AI), to name a few from a lengthy list.

Despite the boom, I anticipate a significant number of these ventures will eventually fold, possibly not in 2024, but soon after. Launching an AI product is one thing; addressing a genuine, valuable need for a substantial customer base is quite another.

Success in the real world, especially in the AI sphere, is challenging. A novel idea isn’t enough if it doesn’t attract a sufficient customer base or offer compelling value. This reality was true before the era of AI and remains so.

Additionally, many AI startups may struggle with defensibility. Without strong protection against imitation, even a successful product can be replicated, sometimes by larger, more established companies, as seen in Apple’s history of integrating features from smaller innovators.

While the future may hold more failures than successes for AI startups, this isn’t pessimism but a recognition of the typical lifecycle of innovation. However, on a positive note, this era will undoubtedly bring a multitude of remarkable AI-enhanced products to the market.

LLMs Evolve into Dependable Tools for Practical Planning

This prediction is somewhat contentious, with debates surrounding both the ‘reliability’ and ‘planning’ aspects of LLMs.

On the topic of reliability, the issue of AI “hallucinations,” where systems fabricate information presented confidently, is well-known. This tendency can mislead users who might mistake the AI’s assertiveness for accuracy.

AI doesn’t inherently understand uncertainty as humans do. Ideally, AI systems would convey an “assurance score” indicating their confidence level, like saying, “I’m not certain, but it seems…” However, no AI chatbot has implemented this feature yet. Incorporating a sense of doubt could enhance AI-human interaction, but I don’t foresee this happening by 2024.

Yet, the accuracy of AI chatbots has been on an upward trajectory, and I anticipate this improvement will continue into 2024. I’ve previously discussed strategies to mitigate hallucinations in AI in my article on stopping hallucinations in various chatbots, before I was aware of the “RAG” (Retrieval-Augmented Generation) technique, which I essentially describe in that piece.

Regarding planning, some AI experts doubt LLMs’ capability in this area, often without substantial evidence. I disagree, believing that LLMs can, and to some extent already do, handle planning tasks effectively. For instance, in my article examining whether Google’s Gemini can plan, I presented promising findings.

I predict that by 2024, we’ll begin to see travel planning assistants and similar tools becoming practically useful, powered by LLMs. Generative AI is advancing steadily rather than in giant leaps, and I expect both reliability and planning capabilities to continue improving throughout the year.

Bonus

The Notion page showcases its platform designed for creating wikis, documents, and managing projects. It features an AI assistant, various templates, and is suitable for teams of all sizes. The platform caters to diverse professional groups and emphasizes community engagement and global events. For more details, you can visit their Notion page.

Thank you for reading this article so far, you can also get the free prompts from here.

https://www.buymeacoffee.com/yukitaylorw

What Will You Get?

• Access to my Premium Prompts Library.

• Access our News Letters to get help along your journey.

• Access to our Upcoming Premium Tools for free.

Also, Check out discounted digital contents on https://www.solan-ai.com/

Subscribe My FREE NewsLetter now!

https://solansync.beehiiv.com/subscribe