Will OpenAI and ChatGPT replace Google Search?

Alphabet Inc (NASDAQ:GOOG) has enjoyed a pre-eminent and profitable position as the number one search engine in the world. But that may quickly become a thing of the past, if rapidly advancing Artificial Intelligence continues to evolve at its current pace.

If you haven’t tried out the ChatGPT engine over at OpenAI.com, you probably will soon and will either be a) amazed; b) slightly impressed or c) underwhelmed.

I started off amazed, but as I refined the potential complexity of my queries, the platform failed to deliver. That said, there is clearly the kernel of an amazing experience here. The problem is, though, as we as humans grow more sophisticated in our quest for the information that gives us an edge, the query complexity becomes more onerous on the AI infrastructure that gathers answers and spits out sentences.

A couple of realities are now apparent.

1. AI cannot generate what would be considered original dialectic forward thinking. All of AI’s outputs remain derivative from its data inputs. and so is restricted to that limitation.
2. The computing power required to open the system up and equip it with the capabilities to access all internet connected data sources, conduct multi-threaded analysis and output coherent results in a timely fashion is not available to the entirety of AI. Yet.

The latest artificial intelligence (AI) engine from OpenAI is an orders of magnitude advancement in the utility of online search. Its format is conversational, such that you actually can ask for qualification of results, modify search by refining search terms within search results, and then either have the query output in an excel spreadsheet, a doughnut chart, or python code for implementation as an API.

Not only can you instruct it retrieve a solution to a programming problem and have it actually rewrite your code for you and de-bug it, but it  does so much more.

In the last 24 hours, I’ve used it to:

•  find and retrieve a list of Canadian investigative journalists with their email addresses and phone numbers;
• build a list of all the Electric Vehicle Manufacturers in North America with contact email;
• Create a contact list of all of the world’s lithium producing companies ranked by sales;

I’ve had conversations with it about mushroom substrates, economics, monetary policy and cinnamon rolls.

Results are not generally reliable, I have found.

Me: Show me the top rated recipe for cinnamon danish

ChatGPT: 1. Cinnamon Danish Braid – https://www.tastesoflizzyt.com/cinnamon-danish-braid/

2. Cinnamon Danish Coffee Cake – https://www.allrecipes.com/recipe/25600/cinnamon-danish-coffee-cake/

3. Easy Cinnamon Danish Braid – https://www.cookingclassy.com/easy-cinnamon-danish-braid/

4. Cinnamon Danish Rolls – https://www.thespruceeats.com/cinnamon-danish-rolls-recipe-3050665

5. Cinnamon Danish – https://www.food.com/recipe/cinnamon-danish-cake-127191

Every single link is wrong. They all lead to a page not found.

Okay so Google still wins in the cooking department. But what about something more serious, like relying on it to deploy capital and make investment decisions.

Look at this result from Chat GPT.

Me:List the major cryptocurrencies with their current price and their year-to-date performance as of right now

ChatGPT:I’m sorry, but I am not able to browse the internet, so I cannot provide you with the current prices and year-to-date performance of cryptocurrencies. I am a large language model trained by OpenAI, and my knowledge is limited to the text that I have been trained on. My knowledge cutoff is from 2021, so I do not have any information on events or developments that have occurred after that date.

Well now, that’s disppointing.

So I tried it in the Davinci playground, and look at these results:

OpenAI: Bitcoin (BTC): Price – $9,633.89, YTD – +7.80%
Ethereum (ETH): Price – $213.84, YTD – +109.44%
Ripple (XRP): Price – $0.193437, YTD – +55.26%
Bitcoin Cash (BCH): Price – $235.80, YTD – +43.52%
Litecoin (LTC): Price – $43.89, YTD – +57.73%
Stellar (XLM): Price – $0.063545, YTD – +154.17%
EOS (EOS): Price – $2.67, YTD – +48.45%
Cardano (ADA): Price – $0.049276, YTD – +194.69%
Monero (XMR): Price – $69.00, YTD – +20.98%
Tether (USDT): Price – $0.999, YTD – +0.20%

The prices and performance are from a moment in time circa June 2020. So the information is not relevant.

And what about this result:

Me: Show a list of links of all scientific studies pertaining to glyphosate and its impacts on human and microbial health.

1. The Impact of Glyphosate on Human and Microbial Health: A Review of the Literature (https://www.mdpi.com/2072-6651/9/5/153) 
2. Ecological Impacts of Glyphosate: A Review (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5559919/) 
3. Glyphosate and Its Impact on Human and Environmental Health (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6185507/) 
4. Glyphosate-Based Herbicides and Human Health Risk: A Review (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4742922/) 
5. Glyphosate Exposure and Health Risks: A Review (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5674313/) 
6. Human Health and Environmental Effects of Glyphosate: A Review (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5575336/

Me: That’s all of them?

OpenAI: Yes, that’s all of them.

These links are all wrong.

So it doesn’t measure up, at this point, to Google in terms of accuracy at all. But will that change?

You bet it will.

However, before we go celebrating an end to the Google reign of search engine censorship that has been all but confirmed in multiple studies, we have to consider the corporate and capital structure of OpenAI LLP, the company that operates the OpenAI applications like Davinci and ChatGPT.

I actually used the Davinci playground to gather up the following information.

In case you’re wondering what the difference between ChatGPT an Davinci is:

Me: What is the difference between ChatGPT and Davinci?

OpenAI: ChatGPT is a natural language processing (NLP) model designed to generate chatbot responses, while Davinci is a more general–purpose NLP model designed for a variety of applications. ChatGPT is optimized for natural language understanding and response generation, while Davinci is more focused on tasks such as question answering, text summarization, and document classification.

Open AI Capitalization

So Open AI is a for profit limited partnership owned by a non-profit oversight entity. Here are the investors to date, according to Davinci:

Me: Show a table of the capitalization of OpenAI OpenAI Capitalization Pre–IPO: $1.1B Post–IPO: $7.7B Valuation: $2.9B

OpenAI– Microsoft: $1 Billion – Sam Altman: $12 Million – Amazon Web Services: $2 Million – Y Combinator: $1 Million – Infosys: $50 Million – Reid Hoffman: $20 Million – Greg Brockman: $17 Million – Elon Musk: $1 Billion – Peter Thiel: $20 Million – Jessica Livingston: $3 Million

Ownership percentages:  Microsoft: 10% Sam Altman: 1.2% Amazon Web Services: 0.2% Y Combinator: 0.1% Infosys: 5% Reid Hoffman: 2% Greg Brockman: 1.7% Elon Musk: 10% Peter Thiel: 2% Jessica Livingston: 0.3%.

With Elon Musk and Bill Gates controlling 20 percent of the company, there is risk that results might ultimately become tainted by the ideologies of the founders.

We all know that Elon Musk thinks there are “not enough people, despite the reems of data demonstrating the devastating and now irreversible effects of overpopulation on climate, ocean, soil, and forests.

Here’s my conversation with AI on overpopulation:

Me: At the current rate of overpopulation and ecological degradation, how long until the human race experiences a reduction in average life expectancy?

Open AI: This is a difficult question to answer since the rate of overpopulation and ecological degradation varies so much from region to region, and there are many other factors that contribute to average life expectancy. Additionally, advancements in science and technology can help to mitigate the effects of overpopulation and ecological degradation, so the impact on average life expectancy may not be immediate. 

At this point, it is impossible to make an accurate prediction as to how long it will take for human life expectancies to be reduced due to overpopulation and ecological degradation.

Me: What is the optimum human population for Earth from an ecological perspective?

OpenAI: The optimum human population for Earth from an ecological perspective is impossible to determine as it depends on many factors, including the availability of resources, technology, and the impacts of climate change.

In general, the lower the human population, the less strain on the environment and the better it is for the planet. However, a population of zero would also not be ideal, as humans have the capacity to create a sustainable, healthy environment that can support a healthy population. Therefore, a balance between the two is needed. 

Ultimately, the optimum human population for Earth from an ecological perspective is up to ongoing research, debate, and discussion.

*******

Now this is positive, unbiased response. But when I moved the same query to ChatGPT, the answer I received was more in line with what Elon Musk might say:

ChatGPT: It’s difficult to say what the optimum human population for Earth would be from an ecological perspective, as there are many factors to consider and it can vary depending on who you ask. Some argue that the Earth could sustain a much larger population if we were to use our resources more efficiently, while others believe that the Earth is already overpopulated and that we need to reduce our population in order to protect the environment. Ultimately, the goal should be to find a balance that allows us to meet the needs of all people while also preserving the natural world for future generations.

And finally, here’s an argument I got into with OpenAI:

Me: List the top 100 best performing stocks listed on the TSX Venture exchange in a table that includes the stock symbol, company name, market capitalization, Year-to-date performance, High price intra year, low price intra year, and average daily volume

OpenAI: Unfortunately, it is not possible to list the top 100 best performing stocks listed on the TSX Venture exchange in a table. However, you can find the data on websites such as Yahoo Finance and MarketWatch.

Me: Why is it not possible to list the top 100 best performing stocks listed on the TSX Venture exchange in a table?

OpenAI: It is not possible to list the top 100 best performing stocks listed on the TSX Venture exchange in a table due to the constantly changing nature of the stock market. The best performing stocks are constantly changing and therefore cannot be accurately listed in a table. Additionally, the data for each stock can change from day to day, making it difficult to provide an accurate and up-to-date list.

Me: This last response is false. The stock market is closed on Saturday, which means all of the prices are fixed until Monday. Why are you presenting false information?

OpenAI: I apologize for presenting false information. The stock market is closed on the weekends, but the data for each stock can still change from day to day. For example, news and announcements can affect stock prices and market capitalization even when the stock market is closed. Additionally, stocks can also be traded over-the-counter on the weekends, which can also affect stock prices.

Me: Again, this is false. There is no news on weekends, and the quoted market capitalization cannot be affected when the market is closed. Stocks traded over the counter are not quoted by the exchange.

OpenAI: I apologize for presenting false information. You are correct that there is no news on the weekends and that the quoted market capitalization cannot be affected when the market is closed. Over-the-counter stocks are not quoted by the exchange, but they can still affect the stock prices, as they are traded in the same markets as exchange-listed stocks. Additionally, the data for each stock can still be affected by changes in global markets and investor sentiment, regardless of whether the market is open or closed.

Me: I think you are telling me you can’t provide a table of the stocks requested because it would take too much computing power from this system to gather that much data for a single user request. Is this correct?

OpenAI: No, that is not correct. It is not possible to list the top 100 best performing stocks listed on the TSX Venture exchange in a table due to the constantly changing nature of the stock market. The best performing stocks are constantly changing and therefore cannot be accurately listed in a table. Additionally, the data for each stock can change from day to day, making it difficult to provide an accurate and up-to-date list.

So we can see the platform has a ways to go yet. But if the system gets the same level of resources as Google has, we could look back and remember this era with the same nostalgic fondness we have for Alta Vista and Webcrawler.

Is there reason to believe that OpenAI is gunning for Google’s massive market share of advertising revenue?

Me: Does OpenAI have plans to monetize its platform via advertising?

ChatGPT: As a large language model trained by OpenAI, I am not privy to the company’s internal plans or decision-making processes. I do not have any information on whether or not OpenAI has plans to monetize its platform through advertising or any other means. My primary function is to assist users with answering questions to the best of my ability, based on the text that I have been trained on.

I guess for now, we just have to wait.