Meta AI and the Chatbot Surge: A Survey
Curator: Parneet Kaur Jasmine Dhamija
Banner by: Siddhant Walia
In our fast-paced world, where technology has infiltrated every aspect of our lives, people yearn not just for solutions to their problems but also for meaningful interactions. From ELIZA’s inauguration in 1966 to modern sophisticated LLM-based Chatbots, we have witnessed massive growth in the amount of personalization the technology could provide. Initially, these simple programs could only understand and respond to basic commands. However, with the advancement in natural language processing (NLP), machine learning (ML) algorithms, and internet access, chatbots can now answer novel problems.
Social media services have capitalized on such desires by leveraging chatbots in their existing interface. With the advent of AI, these digital companions have become even more impressive, giving rise to a huge industry. Recognizing this massive opportunity, Meta integrated Meta AI with Llama 3(now Llama 3.1)—their very own large language model—into Facebook, Instagram, Messenger, and WhatsApp.
This newsletter explores how Meta AI compares to the other AI chatbots in the market and is also based on a live survey that explores people's experiences with these initiatives.
Big Step by Meta
The tech giant Meta has always seized every opportunity to stay ahead of trends and outpace their competitors. On April 28th they decided to focus their efforts on enriching people's lives by embedding Llama-3-powered Meta AI into their services, thereby enhancing user engagement and efficiency.
Building on their commitment, they unveiled Llama 3.1 on July 23rd, which has now also been seamlessly integrated into their services. This is their most advanced AI model to date, boasting a staggering 405 billion parameters, a powerhouse to rival top-tier systems like GPT-4 and Claude 3.5 Sonnet. This game-changing model not only offers developers a cutting-edge, freely accessible tool but has also demonstrated exceptional performance across diverse tasks, from general knowledge and math to multilingual translation. Backed by extensive training on a vast dataset and state-of-the-art technology, this model stands out as one of the most formidable AI systems available. Llama 3.1 offers three impressive variants—8B, 70B, and the flagship 450B—each engineered to push the boundaries of AI capability. Although Llama 3.1 is available for free, it is not fully open source. However, to support its developers, Meta has also introduced a new licensing model that lifts some previous restrictions and permits the use of its outputs for training other models.
This strategic integration of Meta AI as an extension to their existing services including Facebook, Instagram, Messenger, and WhatsApp, could not only enrich the user experiences but also host a testing ground for Meta to experiment with new AI capabilities and features in a more relaxed environment.
Pros and Cons of Messaging Chatbots as Compared to Web Chatbots
Messaging-app chatbots, though extensions of their web-based predecessors offer unique advantages that significantly enhance user experience. Their convenience is unmatched, allowing quick and easy access to AI assistance directly within the apps you use daily. This accessibility broadens the reach of AI-powered help, eliminating the need to install additional apps—a limitation of web-based chatbots. Moreover, their ability to provide contextual support within the conversation flow ensures seamless interactions without the need to switch apps. Additionally, AI assistants on messaging apps inject fun and entertainment into conversations, making them perfect for casual chats and humor.
Despite their many benefits, messaging-app chatbots often fall short in generating long texts or detailed explanations compared to their web-based counterparts. This limitation can result in less personalized responses, as the chatbots may not fully grasp the user's context or preferences. Privacy concerns also arise, with some users hesitant to share personal information or conversations with AI on social media platforms. Additionally, while these chatbots are designed to be empathetic, they may struggle to fully understand the nuances of human emotions, partly due to their limited access to the full capabilities of the underlying language models.
How The Competitors Are Not So Far Behind
Snapchat pioneered the market in the sector of personalized chatbots on April 19, 2023, with their ChatGPT-powered ‘My AI’. The update allowed users to create and name their very own interactive bitmoji assistant. The release garnered a daily turnover of over 2 million chat messages and continues to be popular among its users. The chatbot could not only converse with users, but also summarize their texts, solve their math problems, play text-based games, offer recommendations, and provide directions to locations. As of May 2nd, 2024, the bot also includes a feature for setting up reminders.
Recently at Google I/O 2024, the company announced an expansion of its AI model Gemini 1.5 pro to Google Docs, Google Sheets, Google Slides, and Drive as a chatbot! They have engineered this bot to boost the user experience with features like Quick summarization, Smart Replies, Important Suggestions, and Quick information retrieval. However, this new Gemini chatbot for Google Messages is currently exclusive to smartphones including Pixel 6, Pixel Fold, Samsung Galaxy S22, and Samsung Galaxy Z Flip or Z Fold as well as their later versions. Additionally, the devices eligible for this upgrade are required to have 6GB of RAM or higher.
Has This Tech Become Redundant?
As Numerous companies like Meta, Telegram, and Google are monetizing this trend to its fullest extent, we have witnessed an influx of Chatbots that perform the same task.
This raises a genuine question: "Why do we require so many of these tools if the assistance provided by them is not worlds apart?"
To tackle this question, we queried a large number of students from?various departments across the campus of Chitkara University, and here’s what we got:
Some students expressed skepticism about the proliferation of AI chatbots, seeing them as redundant and unnecessary. Vedansh Vasudeva criticized the trend, stating that companies like Meta are more interested in showcasing their AI capabilities than providing genuinely useful tools. Harshit Thapliyal highlighted the herd mentality among tech companies, pointing out that while the AI boom has led to every company wanting its own AI, this does not necessarily equate to innovation. Gopal echoed this sentiment, arguing that all AI chatbots generally provide the same information and solutions, making the trend repetitive and unnecessary.
On the other hand, several students saw value in the variety of AI chatbots. Jaskaran noted that different chatbots often exhibit distinct personalities and provide varied responses to the same prompts, which could offer a richer user experience. Ayush argued that having AI integrated into different platforms is beneficial, as not everyone uses the same apps, and platform-specific AI could cater to different audiences. Kartik appreciated the convenience of having multiple AI assistants across various apps, as it saves time by reducing the need to navigate between different applications.
The discussion around AI chatbots highlights a divide in opinions, with some viewing the proliferation of these tools as redundant, while others see it as a valuable evolution in technology. The debate reflects the complex nature of technological advancement, where the benefits and drawbacks often depend on individual perspectives and needs.
Public Preferences and Perspectives on Chatbots
Our survey revealed a diverse range of opinions on chatbot preferences, highlighting significant differences across various departments that brought unique perspectives to the table,? prompting us to embrace the various use cases of AI Chatbots.
Insights from the AI Department:
Harshit Thapliyal expressed his preference for well-established tools like ChatGPT over Meta AI, describing Meta AI as a mere novelty. He noted, "If Meta AI can attract an audience, it’s beneficial for the company, but ultimately, the choice should depend on the user’s specific needs."
Dr. Tanvi Sood also shared her experiences with various platforms, including ChatGPT-4, Microsoft Copilot, Gemini AI, and Claude AI. She gave her mark of approval to Claude AI, stating, “Recently, I used this platform called Claude AI for the first time and found its conversational ability to be commendable. If I tell Claude to answer me in a certain manner, it holds on to that command for the entire conversation, and I don’t have to ask it to rephrase over and over again. One drawback, however, is its question limit on the free version.”
Jaskaran expressed a strong preference for ChatGPT-4.0, saying, “Currently, I’m using ChatGPT-4.0 (the paid version), and it stands out for its smooth and intuitive interactions. Compared to Gemini and ChatGPT-3.5, ChatGPT-4.0 feels more personable and reliable. While Snapchat AI and Meta AI on WhatsApp are adequate, GPT-4.0 offers a level of dependability that truly sets it apart.”
Insights from other Departments:
Beyond the AI Department, other perspectives of students like Khushi add to the richness of these insights. Khushi shared her approach to selecting chatbots: “I choose the chatbot based on the task at hand. For content writing, I rely on ChatGPT, while for solving math problems, I prefer Gemini.”
Interestingly, a student from the psychology department at Chitkara University voiced a different perspective. This student believes that AI assistants should remain web-based and views their integration into daily-use apps like WhatsApp or Messenger as unnecessary.
In the design department, Mohak, Abhishek, and Anirudh have distinct preferences. They favor Perplexity and Google Gemini for their specific functionalities. Mohak and his colleagues appreciate Gemini for its design-enhancing suggestions and Perplexity for its succinct answers and helpful bibliography links. Abhishek elaborated, “When we upload a PDF to Perplexity, it delivers relevant case studies and summaries that greatly assist our design research.”
These diverse viewpoints underscore the varied applications and preferences for AI tools, reflecting how different needs drive the adoption of specific technologies.
Other Key Opinions
Our survey also uncovered some intriguing concerns that sparked important discussions about AI integration.
Privacy Concerns with this AI Integration:
Some individuals raised concerns about the privacy risks associated with newly integrated AI platforms. They highlighted that platforms developed by Meta AI, such as WhatsApp and Instagram, involve sharing substantial personal data, which raises privacy issues. One interviewee remarked, "I recommend limiting AI's access to user data to balance functionality with privacy."
Robbing the Soul out of Conversations:
Another perspective shared was, "I don’t think AI-supported systems are required on platforms like WhatsApp and Instagram. With their primary purpose being communication, AI assistants take away the credibility of the conversations and disrupt the natural flow of interactions."
Conclusion
As technology continues to evolve at a breakneck pace, the introduction of AI-powered chatbots across various platforms underscores the relentless drive to enhance user experience and convenience. From Meta's integration of Llama 3 into its social media ecosystem to the innovative features of Snapchat's My AI and Google's Gemini, each platform brings its unique strengths and challenges to the table.
While some argue that the proliferation of these AI assistants may seem redundant, the varied preferences and applications highlighted by users reveal a nuanced landscape. Whether it's for quick queries, personalized assistance, or enhancing productivity, these chatbots offer tailored solutions that cater to diverse needs and preferences.
Ultimately, the surge in AI-driven tools reflects a broader trend toward personalization and efficiency, aiming to simplify interactions and enrich our digital lives. As technology matures, it will undoubtedly continue to shape our interactions, driving us toward an even more interconnected and responsive future.
AI Weekly is a collective effort by EvolveAI, to inform our followers of the constantly evolving world of AI.
Follow us on Instagram to know more about EvolveAI and the events we conduct on a regular basis.
Thank You for reading the sixth edition of AI Weekly, see you next week!
Pursuing Bachelor's in Computer Science specializing in Artificial Intelligence from Chitkara University|| Public Orator @EvolveAI || HTML || CSS || PYTHON || C || C++ ||
7 个月Insightful!
AI&ML Student and Researcher | Business and Entrepreneur enthusiast | B.E. CSE in Artificial Intelligence and Machine Learning | Freelancer
7 个月Well said!