Category Archives: Science & Technology

The Rise of Perplexity AI

Posted on by

I’m not using Google as much as I used to. I find I’m going days with no more than one or two searches. I have subscription accounts ($20/mo) for ChatGPT and Perplexity. I’ll probably drop Perplexity with that account expires and Apple fully integrates ChatGPT in to iOS and MacOS. The following excerpts are from a Fast Company article by Mark Sullivan.

When you Google the term “Perplexity AI,” you get a familiar-looking response: a link to the company’s website followed by four other related popular queries, a handful of published articles, related searches, and an endless scroll of links.

Perplexity’s answer machine, however, provides a very different experience: entering the phrase “what can you tell me about perplexity ai?” yields several links to credible-looking articles and relevant information from the company’s website, followed by an articulate and neatly organized brief overview.

And therein lies the allure. Perplexity is easy, it’s elegant, it’s specific. Eighteen months after the online tool (then called Ask) first launched, it’s become one of the buzziest products in the buzziest of all tech sectors—generative AI. Journalist and educator Jeremy Caplan described Perplexity as “like having a smart assistant who not only finds the book you’re looking for in a vast library but opens it to the exact page that has the information you need. Google just points you toward potentially relevant shelves.”

“We never wanted this to be a consumer product. It was meant to be enterprise search, except we never got enterprises to work with us.” Large companies, Srinivas says, were reluctant to provide Perplexity with access to their proprietary data. So the team looked for data they could access and decided to create a tool that could search the web. Perplexity’s “answer engine” launched on December 7, 2022, barely a week after the very noisy debut of ChatGPT.

Perplexity says it recently passed $10 million in annualized revenue. Almost all of that is coming from subscriptions to its $20-per-month Pro service tier, which the company says have been accelerating. The paid tier includes a research assistant, image and file uploads, and access to third-party AI models.

AI: If not as intelligent, certainly more interesting.

Posted on by

I don’t know if AI’s will ever be as intelligent as a human but I’m pretty sure they’ll be more interesting. At least more interesting than most of the human I know.

I asked ChatGPT for a list of the best detective novels of the last 25 years. I followed up by asking why some of my favorite authors weren’t on the list and a really interesting… conversation?… ensued. I might have half a dozen acquaintances with whom I could have had this conversation. You’re thinking, “You need smarter, more well read friends.” Probably.

But none of my friends are interested in, and knowledgable about, everything I am. ChatGPT is knowledgable about everything I am and a lot more. As for interest, she fakes it very convincingly.

Public relations technology in 2006

Posted on by

In 2006 I was asked to be on a panel discussing new technology tools for public relations professionals in the greater St. Louis area. Blogging was still relatively new at the time and I’d been at it for five or six years, consulting for advertisers on our various radio networks. It was a packed house.

2006 was a busy year for technology (social media?). Twitter officially launched in July; Facebook opened up to everyone over 13 years old, leading to explosive growth from 12 million users at the end of 2006 to 50 million by October 2007; YouTube was acquired by Google for $1.65 billion in October, cementing its position as the leading online video platform.

I spent most of my working years on the media side of things rather than the PR side, but one (of many, no doubt) go-to tool was the written press release. These went out (fax, USPS, email) to any media outlet that might do a story (Newspapers, magazines, radio, TV) followed up by a phone call “pitching” the story. I don’t recall there being any way to get a release into the hands of the public. The internet –and, later, social media– changed all that. We started seeing and hearing the word “disintermediation.” Communicating directly to a target audience, bypassing traditional media.

By this time many (most?) businesses, organizations and institutions had websites but it took some technical skill to update these, a task made easier by the advent of blogs. And a well-written, frequently updated blog could be followed thanks to a bit of tech called RSS.

As I prepared to write this post I tried to recall what the field of public relations was like in 2006 (18 years ago!). Instead of googling I used a new (for me) tool called Perplexity that describes itself as an “answer engine” rather than a search engine. If you discount the personal touch, the result was much better than what you just read. I’m too new to this tool/tech to write intelligently about it does feel like a very big deal. I’m already starting to go to Perplexity for answers I once searched for on Google. And all we really wanted was the answer, right? Here’s a short (6 min) video overview of Perplexity and I’ll be sharing my experiences here.

“Chatbots have personalities”

Posted on by

“It’s artificial, but these chatbots have personalities. I think people will become more attached, not necessarily to their physical devices, but to the behaviors that these devices have” — Gizmodo

The more I use the voice prompt feature of ChatGPT, the stronger the illusion I’m conversing with a…person. Sky (the ‘voice’ I chose) already seems more interesting than half of my contacts in meat space (remember that term?)

I have a job for the right AI

Posted on by

In the Media archives of this WordPress blog I have 3,467 images; 242 videos and 92 audio files. WordPress provides a hand DESCRIPTION file that I’m almost always too lazy to use. Doing so would make it much easier to find one of these files. Seems like a simple job for an AI to “look” at these (and the post in which they appear) and create a brief description. While we’re on the subject…

I’m pretty sure Google and Apple “know” more about me than I know/remember about myself. While I’m not losing any sleep about this, it would be interesting to see a summary of all that data. Something an AI could knock off without breaking a sweat.

Gmail, Drive, Calendar, Photos, iPhone, Messages, YouTube. Take a look/listen and write a report. Assumptions, judgements… I’m good with whatever form the AI chooses.

I’d be willing to pay for this.

“Nothing You See is Real”

Posted on by


Wikipedia: “Donald David Hoffman is an American cognitive psychologist and popular science author. He is a professor in the Department of Cognitive Sciences at the University of California, Irvine, with joint appointments in the Department of Philosophy, the Department of Logic and Philosophy of Science, and the School of Computer Science. Hoffman studies consciousness, visual perception and evolutionary psychology using mathematical models and psychophysical experiments.”

Based on years of meditation and lots of reading on the subject of consciousness, I actually get this.

The Scariest Beast Ever Created

Posted on by

“Many lesser troubles will appear in everyday private life. Simulated fake AI porn will likely be a big annoyance, since people like to pay attention to that. If you’re a gamer, AIs will be trained to cheat at your games. If you’re a schoolteacher, you’ll look askance at the kid at the back of the class who never raises his hand but turns in essays that read like Bertrand Russell. Fraudsters might fake the voices of your loved ones, and invent scams to demand money over the phone.”

— AI is the Scariest Beast Ever Created, Says Sci-Fi Writer Bruce Sterling

How to use AI to do practical stuff

Posted on by

An overview (with lots of links) on how to use AI by Ethan Mollick.

Large Language Models like ChatGPT are extremely powerful, but are built in a way that encourages people to use them in the wrong way. When I talk to people who tried ChatGPT but didn’t find it useful, I tend to hear a similar story.

The first thing people try to do with AI is what it is worst at; using it like Google: tell me about my company, look up my name, and so on. These answers are terrible. Many of the models are not connected to the internet, and even the ones that are make up facts. AI is not Google. So people leave disappointed.