Thank you for such a good writing. I am a commercial researcher and I do have the same feeling with Perplexity. However, there's one thing I have been trying to crack, Perplexity get any sources available on the internet (blog post, youtube videos, whitepaper or any public pdf files), as long as it returns an answer to our prompt.
Thank you for such a good writing. I am a commercial researcher and I do have the same feeling with Perplexity. However, there's one thing I have been trying to crack, Perplexity get any sources available on the internet (blog post, youtube videos, whitepaper or any public pdf files), as long as it returns an answer to our prompt.
This lead me to next question: how can we reduce/increase the quality input when using Perplexity? For this reason, I have also been using Bing and it seems that Bing returns with (skewed towards) whitepaper/reports and articles as their sources. Certainly I know I am using both free options for these tools, so I'm unsure what it gives in paid option.
In short, what is your take on the quality of source in doing market research with AI?
Perplexity provides you with the all of the sources its using in its answer/output. The best thing we can do at this point is take a look at each of those sources. Make sure they are reliable and not some random unknown blog or something.
You can then deselect those sources from Perplexity and it will adjust its response in real time.
Still takes a little human oversight, but that's still where we are at with AI. And it's still way faster and better than Google most of the time.
You hit the nail on the head. The quality of the sources is the main problem with Perplexity-for both the free and paid version. I use Perplexity a lot and became a bigger fan after listening to a recent Lex Fridman podcast with Perplexity's CEO. I wrote a recent article about Perplexity and my qualms with the quality of its sources, which included reddits and even one of my own LinkedIn articles.
Perplexity has been sued by various media outlets, which has further left them scouring the internet for credible sources.
I short, Perplexity is a great starting place to get ideas and knowledge, but for anything resembling high stakes research, you still need Google whose algorithm vets more credible sources, and even better, Google Scholar.
I'm glad to read this tbh, because that is what I have been doing too (and wonder if I am too old-fashioned in this manner).
I do use AI, any tools, Perplexity/Bing or GPT 4.o to have an idea of what the trends/summary might be. And then I go back to Google the source that it gives and find back-up credible evidence for them. Thank you Nigel.
I see that you are promoting your work in my comments section Nigel. I think you might find that this is not the most effective way of doing marketing or building relationships. Very few people will ever visit the comments section and it might annoy the Creators of the articles you are attempting to hijack for your own benefit.
Probably the best way to supercharge your new publication is by doing guest posts with others on their Newsletters that are aligned to the topics you care about and can provide the most value. At least in my experience, those are the cross-promotions that convert the highest quality new readers.
The recommendations and Notes driven new readers aren’t necessarily the highest native quality since the topics that we write about are fairly “new” to the baseline audience in this ecosystem.
Thank you for such a good writing. I am a commercial researcher and I do have the same feeling with Perplexity. However, there's one thing I have been trying to crack, Perplexity get any sources available on the internet (blog post, youtube videos, whitepaper or any public pdf files), as long as it returns an answer to our prompt.
This lead me to next question: how can we reduce/increase the quality input when using Perplexity? For this reason, I have also been using Bing and it seems that Bing returns with (skewed towards) whitepaper/reports and articles as their sources. Certainly I know I am using both free options for these tools, so I'm unsure what it gives in paid option.
In short, what is your take on the quality of source in doing market research with AI?
You can use the “academic” focus mode to have Perplexity only search through academic papers.
Perplexity provides you with the all of the sources its using in its answer/output. The best thing we can do at this point is take a look at each of those sources. Make sure they are reliable and not some random unknown blog or something.
You can then deselect those sources from Perplexity and it will adjust its response in real time.
Still takes a little human oversight, but that's still where we are at with AI. And it's still way faster and better than Google most of the time.
For very quick searches I'm finding genspark dot AI to be somewhat useful. The citations are clearly placed.
You hit the nail on the head. The quality of the sources is the main problem with Perplexity-for both the free and paid version. I use Perplexity a lot and became a bigger fan after listening to a recent Lex Fridman podcast with Perplexity's CEO. I wrote a recent article about Perplexity and my qualms with the quality of its sources, which included reddits and even one of my own LinkedIn articles.
Perplexity has been sued by various media outlets, which has further left them scouring the internet for credible sources.
I short, Perplexity is a great starting place to get ideas and knowledge, but for anything resembling high stakes research, you still need Google whose algorithm vets more credible sources, and even better, Google Scholar.
I'm glad to read this tbh, because that is what I have been doing too (and wonder if I am too old-fashioned in this manner).
I do use AI, any tools, Perplexity/Bing or GPT 4.o to have an idea of what the trends/summary might be. And then I go back to Google the source that it gives and find back-up credible evidence for them. Thank you Nigel.
You're welcome, Norah. The post I alluded to above was a LinkedIn post in which I emphasized what I take to be the main purpose of tools like Perplexity and Claude: to be curiosity drivers (FYR: https://www.linkedin.com/posts/nigeldaly_ai-genai-ethicalai-activity-7211174939737948160-8puI?utm_source=share&utm_medium=member_desktop).
I see that you are promoting your work in my comments section Nigel. I think you might find that this is not the most effective way of doing marketing or building relationships. Very few people will ever visit the comments section and it might annoy the Creators of the articles you are attempting to hijack for your own benefit.
Point taken and apologies if I overstepped bounds.
Probably the best way to supercharge your new publication is by doing guest posts with others on their Newsletters that are aligned to the topics you care about and can provide the most value. At least in my experience, those are the cross-promotions that convert the highest quality new readers.
The recommendations and Notes driven new readers aren’t necessarily the highest native quality since the topics that we write about are fairly “new” to the baseline audience in this ecosystem.