Google Joins the Race

Google officially announced their ChatGPT competitor (named Bard A.I.) today. Many people have known that Google has been investing significantly in AI for many years now, much enhanced by their acquisition of Deepmind in 2014. I will be eagerly awaiting their release, which will initially come out as a 'light-weight' version to help scale to the millions of users that will likely test out the product on its debut.

We've made significant strides in A.I. in the last twelve months, with companies like OpenAI, Stability AI, Lensa, Google, and more making breakthroughs and headlines across many different industries. As the access to 'AI and AI tools' becomes easier and cheaper in the coming months and years, the main thing I'd like to think about is what companies and people will be doing to adapt to this.

What to Expect in the Coming Years for AI

Over the last few months, I see a lot of startups trying to hop onto the 'AI train' and try to incorporate AI as a feature or product to their underlying business, yet this novelty will soon wear off as the access to AI will become as omnipresent as using an online search engine like Google, Bing, Yahoo, DuckDuckGo, or anything else. In reality, a lot of companies are offering AI by just implementing an API from companies like OpenAI and trying to rebrand themselves now as an "AI company. Well as I'd like to say: "You're not an AI company because you added the OpenAI API code into your product, just as you're not a fintech company for adding the Stripe API into your checkout pages."

A more important skill in the coming future will be the ability to speak to AI. Some companies are already starting to hire for positions like these, calling this newly created position an "AI prompt engineer" for now. Since access to AI is no longer the bottleneck, one key differentiator that will arise is how you interact with the AI. The best companies will focus more and more on building a collection of prompts that are the most useful for their use cases, persona, and business. This will require a lot of trial and error, along with a good understanding of how the business works, and how the AI model understands human speech.

Data will be another core differentiator for businesses going forward. If everyone is utilizing the same AI models that were trained on the same data, things will quickly start looking and sounding the same. To prevent this, I foresee a lot of companies building a proprietary database of information that sets their AI models apart from others. This can potentially give better insights that no one else is able to get without the same data, or generate more rich content for their users.

However things play out in the coming future, I will be interested in seeing how companies, organizations, governments, and people evolve to deal with the oncoming advances in AI across all industries.

Initial Thoughts

As a personal side project, I'm also going to be building a social media analytics tool at the same time (CloutSense.io), that utilizes a mixture of Bubble.io, Python, Javascript, Postgresql, and APIs. It's funny, because initially I just wanted to check out Bubble.io as a no-code platform and try to build something without any code, but I ran into quite a few limitations and ended up needing to support some parts of the project with actual code, forcing me to brush up more on Python and Javascript. After spending some time on it, I've realized that Bubble does help accelerate the process for someone trying to make a prototype or MVP to test product idea online, but it falls short in a few areas, particularly in the backend.

Current Progress

I've gotten the base dashboard setup so that it can display the main information to track the changes in social influence. The two social channels that I've decided to work on first were Twitter and LinkedIn. Twitter because it is more developer-friendly and easier build a tracker for than other channels & LinkedIn because I personally would love to have a dashboard to track changes in employee size for various portfolio companies and other startups that I'd like to monitor for the long run. Here's an image of what the LinkedIn dashboard currently looks like for my internal dashboard:

What's Next

I've made a good amount of progress on the base layer of my app and have been collecting a decent amount of data to get started. For what's next, I'll be working on ironing out the initial onboarding UX and implementing a basic upgrade flow for users to increase their account tier (and usage limits). I would love to launch an MVP version of CloutSense by early Q1 2023 for the social channels Twitter and LinkedIn. Depending on if the initial platform gets any traction or not, I'll decide whether to focus on expanding the social channels supported or tweaking the product features itself to see there could be more product-market fit.

In a combined effort to start writing more and learn more about no-code platforms like Webflow, I've decided to launch a website from it and start publishing a few blogs here and there about anything I'm working on or thinking about. The topics I'll write about can be anything, but these days it'll be more focused around technology, climate tech, product management, growth, and angel investing.