Cracking the Instagram Code: How ElevenLabs helped me to go ‘Sort of’ Viral

13

October

2023

5/5 (2)

In the formative years of social media, achieving virality was generally limited to a select group of influencers who had built a sizeable following through years of dedicated effort. However, the emergence of short-form content has significantly decreased the threshold for going viral and reaching a broader audience. Due to the inherent properties of short-form content, platforms are incentivized to boost more posts, thereby enabling more creators to gain exposure. However, this also means that the attention span of users on such platforms to engage with content has decreased significantly as well (TLMUN Herald, 2023). So, although the threshold to go viral has been lowered, the challenge of resonating with users and retaining a dedicated audience has become more difficult. 

I personally noticed this trend and set out a goal to create 3-4 reels on Instagram each week to crack the virality code. A prominent tactic for up-and-coming creators on Instagram is to combine their videos with a motivational audio track featuring the voice of a popular influencer. Unfortunately, many of these audio samples are oversaturated. Moreover, my personal voice lacks the booming and motivational attribute to effectively hook doom scrollers. 

As a result, I turned to ElevenLabs, a web-based application enabling users to produce lifelike speech through its AI-assisted text-to-speech software. The platform provides pre-designed voice profiles that realistically synthesize vocal emotion and intonation. In my post, I decided to select the voice of Adam, characterized on the platform as a deep-voiced American narrator (ElevenLabs, 2023). Here’s the input that I provided to the tool:

“If you say you’re gonna wake up at 4:00 in the morning to work on your goals. You wake up at 04:00. It’s going to feel uncomfortable. It’s not going to be fun. Do something that sucks every single day of your life”

If you’re a personal development junkie, you might notice that this text is a direct quote from David Goggins. However, instead of using Goggins’ voice, I opted for Adam’s from ElevenLabs. I shared this reel on September 14th, and the results were surprising: 12,219 impressions, 12,097 unique accounts reached, 1,062 likes, and 310 saves. You can check out the post yourself by clicking on the following link:

The advent of AI tools provides us with innovative approaches to content creation and can often lead to surprising outcomes. You should try it out for yourself!

Bibliography

ElevenLabs. (2023). Overview. Retrieved from ElevenLabs: https://docs.elevenlabs.io/voice-library/overview

TLMUN Herald. (2023, May 24). The Rise of Short Form Content. Retrieved from Medium: https://medium.com/tlmun-herald/the-rise-of-short-form-content-bbd7a47b0d6d

Please rate this

Beyond the Buzz: ChatGPT and the Issue of Input

4

October

2023

No ratings yet.
Revolutionizing AI Art with Chat-GPT 4's Advanced Conversational Capabilities | by Mr Momoh | Bootcamp

The potential impact of AI and its practical implications for many aspects of our daily lives have been widely communicated to us in the past. However, most of us didn’t have the opportunity to gain firsthand experience of how disruptive and powerful this technology could be. So, when credible people like Stephen Hawking and Elon Musk warned us about the potential dangers of AI, it often felt like a distant concern, and many of us didn’t take it seriously (Kharpal, 2017). This sentiment, however, changed rapidly when ChatGPT made its first introduction into our lives. For most, ChatGPT’s responsiveness and general capabilities posed an unprecedented experience with a machine, leaving many of us with an existential fear of being replaced in the future (Brower, 2023) . To rationally assess such concerns, it’s essential to delve into the intrinsic properties of the algorithms underpinning AI applications like ChatGPT.

A brief explanation of its algorithms

In its formative years, AI applications primarily relied on supervised and unsupervised learning algorithms. These algorithms were inherently reliant on structured data to process information and solve data-intensive problems (Kar, 2016). In contrast, ChatGPT harnesses the use of deep- and reinforcement learning to mirror human intelligence. Consequently, GPT-4 can process and, in turn, produce unstructured data such as text, images, audio, and video (Dwivedi et al., 2023). Due to its intricate nature, however, ChatGPT functions as a black box, i.e., it remains unclear how it arrives at its output based on the input given to its algorithms. What is evident, however, is that its reinforcement protocols continuously improve through a constant influx of new training data. So, in fact, ChatGPT mostly produces recycled original content presented in a coherent and logical manner (Chandra et al., 2022). 

The issue at hand

So, what precisely is the downside of this? The issue at hand revolves around the surge of new and highly capable AI applications, which has fostered a sentiment that humans are replaceable. A great example of this trend can be found in the art world. Since the advent of GPT-4 and Dall-E, many artists have become discouraged by the depreciating effect of these technologies on the value of artwork (Ciftci, 2023). The irony in this lies in the fact that these applications inherently depend on human-generated original input. This begs the question of how these applications will evolve as humans generate less original content over time. 

So, are we spearheaded towards a future burdened by a lack of original content, or will the situation stabilize itself as people come to realize that these sophisticated algorithms are essentially advanced recycling systems? Regardless of the outcome, I firmly believe in advocating for a future that stimulates an environment where individuals are incentivised to produce original content. Let’s not compromise the creative capabilities of our future generation by displacing it with unoriginal, mass-information-inducing algorithms.

Bibliography

Brower, T. (2023, March 5). People Fear Being Replaced By AI And ChatGPT: 3 Ways To Lead Well Amidst Anxiety. Retrieved from Forbes: https://www.forbes.com/sites/tracybrower/2023/03/05/people-fear-being-replaced-by-ai-and-chatgpt-3-ways-to-lead-well-amidst-anxiety/?sh=4921043d7fe6

Ciftci, B. (2023, March 12). Why some artists see AI-generated art as a threat to their livelihood. Retrieved from Medium: https://bootcamp.uxdesign.cc/why-some-artists-see-ai-generated-art-as-a-threat-to-their-livelihood-f4634b24a5ce

Chandra, S., Shirish, A., & Srivastava, S. C. (2022). To Be or Not to Be …Human? Theorizing the Role 

of Human-Like Competencies in Conversational Artificial Intelligence Agents. Journal of Management Information Systems, 39(4), 969–1005.

Dwivedi, Y. K., Kshetri, N., Hughes, L., Slade, E. L., Jeyaraj, A., Kar, A. K., Baabdullah, A. M., 

Koohang, A., Raghavan, V., Ahuja, M., Albanna, H., Albashrawi, M. A., Al-Busaidi, A. S., Balakrishnan, J., Barlette, Y., Basu, S., Bose, I., Brooks, L., Buhalis, D., & Carter, L. (2023). “So what if ChatGPT wrote it?” Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy. International Journal of Information Management, 71, 102642. https://doi.org/10.1016/j.ijinfomgt.2023.102642

Kar, A. K. (2016). Bio inspired computing–a review of algorithms and scope of applications. Expert 

Systems with Applications, 59, 20–32.  https://doi.org/10.1016/j.eswa.2016.04.018

Kharpal, A. (2017, November 6). Stephen Hawking says A.I. could be ‘worst event in the history of our civilization’. Retrieved from CNBC: https://www.cnbc.com/2017/11/06/stephen-hawking-ai-could-be-worst-event-in-civilization.html#

Please rate this