Are We At Peak AI? (No.)

Paul Yang - July 17th, 2023

The generative AI craze dominates today (July 2023) across all high-tech companies, whether sector news, earnings reports, or product planning sessions. But much of the joking about how frequently CEOs now say “AI” is anecdotal. We wanted to understand just how pervasive this trend is, and possibly understand a little bit more about who actually is talking about AI.

We took a look at the NASDAQ 100 in conjunction with 189 additional software/technology companies to form a broad convenience sample of high innovation, high technology companies generally. Specifically, we asked “how did their homepages change from January to July 2023?” We combined research done with OpenAI, web scraping through Python, and visualization to reach our conclusions. (And yes, we remembered to only match “AI” the cased whole word.)

In short, our high-level findings confirm an intensification in mentioning “AI,” but the hype has not yet infected the majority of high-tech companies.

Results

Mentioning AI

The percentage of companies mentioning "AI" did not significantly increase. In January 2023, 28% of Technology companies mentioned AI, and this figure only slightly grew to 36% by July. A small percentage of non-technology companies also mentioned AI, but the increase was only noticeable by June.

At this point, only about 40% of software companies (the highest meaningful subgroup) mention AI, indicating that the trend has not spread to all companies. While it's likely that AI can benefit almost every company, it will be interesting to observe the battle between hype and cynicism as it plays out.

Most frequent mentions of AI: A few companies that have chosen to most prominently feature AI include LivePerson (32 mentions!), Sprinklr (17 mentions), and Nice (14 mentions). (Click the link for a full list)

However, more than half of companies who did mention AI in July (55 / 92) had increased their frequency of mentioning “AI” since January 2023. The total number of mentions more than doubled across all companies overall, showing an intensification.

Biggest increase in mentioning AI: A few companies that increased how prominently they featured AI include LivePerson (+15 mentions), Sprinklr / Zoom (+11 mentions), UiPath / Fiscal Note (+10 mentions).

Mentioning Generative AI

More tellingly, for the term “Generative AI,” 0 companies mentioned it directly in January, 2 in March, and 13 by July 2023.

We were somewhat surprised to see companies eschewing the more descriptive version of AI. For a mass audience who maybe did not already follow the AI / ML space closely, it’s possible that “AI” has become synonymous with “Generative AI” and also “ChatGPT.” Conversely, a cynical analysis might suggest that there is an intentional conflation of the two concepts (GenAI vs. AI) so that more companies can discuss their “AI” technology without actually implementing the flavor of the month in LLMs and other generative tools.

Who mentions Generative AI, not just AI: A few companies that have chosen to prominently feature Generative AI include LivePerson, Zscaler, CrowdStrike, and PegaSystems.

So that made us curious about how many companies mentioned “GPT” as the branded, but more commonly colloquial equivalent to generative AI. It turns out, not too many use the branded term (not sure if it’s trademarked since it’s research work!). It ranges from classes about ChatGPT, to mentioning GPT models, to more SEO/clickbait. In particular, Fair Isaac (a.k.a FICO) is running the headline “Is ChatGPT the Weight Loss Wonder Drug of AI?” trying to capture both Ozempic and ChatGPT traffic to its consumer credit risk products, I suppose.

See Also: Betteridge’s Law of Headlines

Who mentions GPT, not just Generative AI: A few companies that have chosen to prominently feature GPT phrasing directly are Udemy, Freshworks, Coursera, Fair Isaac, Sprinklr, and Upland.

Data Gathering Methodology

We proceeded with two broad steps;

  • Use OpenAI services to augment a list of companies with industry and URL
  • Use BeautifulSoup in conjunction with requests to the Wayback Machine in order to see the content of webpages in the past

Generating the Root Data

We started with a list of companies on the NASDAQ 100, and an additional 189 publicly traded companies in the software industry. We then asked GPT3.5 to generate a series of Industry/Sub-Industry classifications and provide the companies' homepages. We wanted the industry label, because many of the companies listed might not be technology companies (Mondelez) but didn’t want to spend too much time manually sifting through lists.

This showcases the power of Generative AI technologies to create root datasets for analysis quickly. This GPT-generated original dataset can be found in this Einblick table.

Grabbing Historical Web Data

At the Wayback Machine, historical copies of websites are stored. Snapshots are taken on a regular basis. For large, publicly traded companies this is a frequent snapshotting process.

We can loop through and find the URL by date:

request_url = f"https://archive.org/wayback/available?url={url}&timestamp={date}"

And then use BeautifulSoup to parse the result:

soup = BeautifulSoup(snapshot_response.content, 'html.parser')

And glue all the dates and URLs together:

results_df = results_df.append({'URL': url, 'date': date, 'hasAI': has_ai, 'hasGenAI': has_gen_ai,
'hasGPT': has_gpt, 'countAI': count_ai, 'countGenAI': count_gen_ai,
'countGPT': count_gpt}, ignore_index=True)

You can see the full data-grabbing code in this Einblick cell.

Cloneable Analysis and Data

If you're interested in exploring further, feel free to clone the data and analysis. Simply open and fork the canvas below.

About

Einblick is an AI-native data science platform that provides data teams with an agile workflow to swiftly explore data, build predictive models, and deploy data apps. Founded in 2020, Einblick was developed based on six years of research at MIT and Brown University. Einblick is funded by Amplify Partners, Flybridge, Samsung Next, Dell Technologies Capital, and Intel Capital. For more information, please visit www.einblick.ai and follow us on LinkedIn and Twitter.

Start using Einblick

Pull all your data sources together, and build actionable insights on a single unified platform.

  • All connectors
  • Unlimited teammates
  • All operators