ChatGPT, feedback from marketers, community managers and … top data scientists

————- This article is fully written by a human … but the anchor image was generated by an AI ——————-

Unless you have been living under a rock, you have heard of the new ChatGPT assistant made by OpenAI.

ChatGPT (as described by ChatGPT)

Using eCairn, we analyzed the buzz around ChatGPT with the objective to see and contrast how people from different communities were talking about it and to identify interesting articles for those who want to explore.

We zoomed on three communities and groups of influencers that we monitor on a daily basis: CMOs, Community Managers and Data Scientists.

Here is a detailed view about the perception of ChatGPT in these three communities:

CMOs and Marketing Execs

We see that the 6000+ CMOs and Marketing Execs we monitor only tweeted 191 times about ChatGPT during the last month. This is very little i.e 0.03 tweets per person.

Here is what CMOS saw the most in ChatGPT:

Clearly the business performance was the key element resonating with CMOs and marketing executives, along with stories about “replacing humans”.

• The business performance of 1M users in 5 days was what they were the most impressed with: https://twitter.com/wesnichols/status/1600223007674224640 from @GRDecter

• Sam Altman’s tweet was also shared a lot: https://twitter.com/sama/status/1601731295792414720 . I encourage you to read the thread. Some people are clearly very skeptical about the real usefulness of ChatGPT.

• On a personal note, I  really liked Greg’s Verdino tweet: https://twitter.com/gregverdino/status/1603129603349938182

Community Managers:

The ~20 000 community managers and developer relation managers we are listening to tweeted 3500 times about ChatGPT in the last month. This is 0.175 Tweets per person so 60 times more than CMOs and Execs.

What do community managers and devrel people say about ChatGPT? Keep in mind that a lot of these people are tech/engineers.

Most of the tweets are describing use cases and provide examples of what you can do with ChatGPT: create content, create code, answer questions, and replace “human tasks”. The chatter is more “neutral”

Here are the articles that these people shared the most (except the ChatGPT login):

Community managers talked about the stack overflow ban,  sharegpt ( an interesting service collecting ChatGPT Q&As) and a great introduction from @swyx

• https://www.theverge.com/2022/12/5/23493932/chatgpt-ai-generated-answers-temporarily-banned-stack-overflow-llms-dangers 

• https://simonwillison.net/2022/Dec/5/rust-chatgpt-copilot/

• https://sharegpt.com/

• https://lspace.swyx.io/p/everything-we-know-about-chatgpt

Data Scientists

As expected, the data scientist influencers we are listening to ( 3000 of them) are the most vocal about chatGPT with 3200+ tweets in the last month i.e more than 1 tweet per person. They also were the fastest to talk about it with significant # of tweets on Nov 30th

What did they have to say about it?

Most shared articles are general presentation of the capabilities and are overall positive:

• https://arstechnica.com/information-technology/2022/12/openai-invites-everyone-to-test-new-ai-powered-chatbot-with-amusing-results/

• https://www.theverge.com/2022/12/8/23499728/ai-capability-accessibility-chatgpt-stable-diffusion-commercialization

• https://www.theguardian.com/technology/2022/dec/04/ai-bot-chatgpt-stuns-academics-with-essay-writing-skills-and-usability

• https://thefinancialbrand.com/news/data-analytics-banking/artificial-intelligence-banking/interview-with-chatgpt-on-the-future-of-ai-assistants-in-banking-156941/

• https://mentalcontractions.substack.com/p/openais-chatgpt , enthusiastic from Pawel Pachniewski

I also discovered thru this listening, OpenAI fake detector which is pretty neat: https://huggingface.co/openai-detector

Digging into the negative keywords ( wrong OR problem along with chatGPT), we find articles that point to the limits of chatGPT .

Here are the most popular ones:

• This MIT article was shared by many data scientists: https://www.technologyreview.com/2022/11/30/1063878/openai-still-fixing-gpt3-ai-large-language-model

• This amazing article, starting with a wrong response from chatGPT, explains the hows and the whys: https://stratechery.com/2022/ai-homework/ ( it takes a good 30 minutes to read)

• This article: https://hai.stanford.edu/news/moderate-proposal-radically-better-ai-powered-web-search offers alternative that blend search and q&a models.

Other articles cited with a few mentions highlight other problems with chatGPT:

• https://www.spectator.co.uk/article/chatgpt-a-world-class-bs-machine/

• https://www.latimes.com/business/story/2022-12-09/a-chat-with-chatgpt

• https://www.forbes.com/sites/lanceeliot/2022/12/13/digging-into-the-buzz-and-fanfare-over-generative-ai-chatgpt-including-looming-ai-ethics-and-ai-law-considerations/?sh=25b27348600c

As a summary, first impression from most people was really positive and ChatGPT is clearly a major improvement comparing to chat solutions that have been around so far. But…there is still a long way to go and (IMHO) the issue of trustworthiness is far from been resolved.

If you’re interested in looking at how people in your target audience/ community talked about ChatGPT, just ask using the contact form