7000 Subreddits go dark and nobody notices

By the time you are reading this, last week's effort of subreddits (imagine shareable links that can be commented on by many within a Reddit community) going private where only the Mods can see content and its subscribers can neither see nor access the subreddit). 

Reddit changed its API usage policy so that it now charges $0.24 per 1,000 API calls; whereas for the past x-number of years API usage was free to developers. Most for-profit companies will charge for API calls; sometimes this is baked into the service contract for enterprise platform licensing and sometimes API usage is charged at the per user per month/year/time period rate. It is not unusual to see online platforms doing this. Even Google charges for its Maps API.

Apollo is a freemium app; meaning that sure, there are features of the free version of it. But, to get access to other features, Apollo Pro costs a flat fee of $4.99 and Apollo Ultra has a subscription cost of $1.49/month or $12.99/year. The app's paid features allow users to manage multiple accounts (Reddit's app only allows one user account per current login session). Apollo has 1.3 million to 1.5 million monthly active users, and of that 900,000 daily active users. Each Apollo user consumes about 10,000 API calls per month to Reddit servers. The Apollo app makes about 50 million API calls per month.

App developers like to point out that Imgur's API pricing is much lower; but they fail to also disclose that Imgur is a free image hosting app. Imgur isn't charging customers to use their services for image sharing because the site is ad supported. 

Here's a thought. Why should Reddit shoulder the maintenance and server costs for all the API calls made by 3rd party vendors?

This Summer's Podcast List

Podcasts are almost like listening to 30 minute to an hour segments of an audiobook, except narrated with unscripted speakers; usually. You'd be surprised how much of what you might think of as an ordinary podcast done by an individual or a couple aren't scripted but they are. The average solo shop Youtube content creator will say that it takes roughly 20 hours of video and sound editing time for every one hour of finished content. Ads pays podcasters roughly $10/thousand listens for short ads, up to $25/thousand listens for longer ads. Let's say that you're Ira Glass of This American Life with 2 million listens per weekly episode.

2 million / 1000 = 2000 x ($10 or $25) = $20,000 to $50,000 potential ad revenue per episode

Influencer Marketing website has an interesting calculator of how profitable podcasts are. This year (2023), podcasts will exceed $2 billion in ad revenue.

But after deducting all the production costs, it still makes more sense for Podcasts to feature many ads. New podcasts are likely to have no ads or start with a couple really short ads at the beginning of the podcast. More established podcasts seem to have ads at the beginning, middle, and end of the podcast. But, how many ads should you serve up to listeners and is it more about greed than content? If you serve up so many ads that you lose listeners. Even though the Apple podcast app has a feature that lets you skip ahead (or "rewind" a few seconds), some podcasts have 4-6 minutes of ads.

When you think of the spoken word, it's roughly 1 page of scripted content per minute or 500 words per page. If you have unique and original podcast content that's 10 minutes long (of just content, no ads), that's 10 pages of words that a human needs to write for the podcast. There is just something about human-written content that is more impactful, meaningful, or entertaining.

I tried to get Bing AI to imagine itself as a shape

Prompt: If if you could imagine yourself, bing chat, in an artificial body, what shape would that body be?

Bing AI: I’ll try to create that.

And then, I got this image as a response...

Looks like self realization is one of those blocked keywords.

Netflix and Chasing Customer Share of Wallet

Here is a rudimentary definition of share of wallet:

Money spent on a product divided by a total amount spent in that product category

Let's say that a person consumes media with the following channels:

• AM/FM Radio (free)
• Satellite radio $16.99/mo - $21.99/mo via SiriusXM
• Podcast subscriptions ($4-5.49/podcast/month)
• Audio books subscription $7.95-$14.95/mo via Audible
• Television (basic broadcast channels are mostly free with ads)
• Satellite TV - DIRECTV ($64.99–$154.99/mo), DISH ($79.99–$109.99/mo)
• Amazon Prime Video (14.99/mo or $139/yr)
• Print newspaper, $520/yr NY Times (full experience, daily print Mo-Su); sad, LA Times does not have this service for my zip code; local newspaper daily digital + print Sunday ($25/mo); the other local newspaper $21-26/mo daily digital + print Sunday
• Print magazine, avg $20/magazine/year (excluding luxury magazines)

The affluent consumer likely subscribes to more than one content service and probably spends more than $500/year on the media-as-entertainment category. Amazon Prime Video's wallet share of the above scenario is 27.8%. It's really hard for other streaming services to compete with Amazon's business model, since Prime Video is a feature of Amazon's Prime service not its core reason for existing. Everyone else is just a one-trick pony.

Consumers are subscribed to multiple services. The average US consumer subscribes to 12 paid subscriptions spread across videos, music, gaming, news, food delivery, and other entertainment. Netflix raising its prices shouldn't be that big of an issue. Is it?

[insert chat AI name], what are today's top headline news stories

It would appear that Bard has a lot of trouble with:

• determining what today's (insert date) headlines are

• getting headlines from it's own news source, Google News

• relevant content when filters are requested (e.g., no politics, no war, no gun violence, no Trump, and no Congress)

When I asked Bard if it has access to Google News, it said it did. Then I asked it to get me a list of the top stories from Google News, but it just reiterated the news stories when I asked for news stories of the day.

Bard at least complied with the filters and did not give me any news stories that had vague interpretation matches to those topics.

The problem, IMHO, lies in that Bard does not cite its sources nor provide links to where it got its answers from; unlike Chat-GPT and Bing AI. 

For top news stories, Bard is not replacing Alexa any time soon. Bing AI is capable of performing this feat with link and source citations, its responses lead you to links that are chock full of ads. 

I'm not impressed by any chatbot capability and what Bard has isn't artificial intelligence. Today, this is no more than a fancy rebranding of machine learning. How can a machine learn on its own if it has to be told to learn on its own and told what sources to use from the Internet. Every response seems to have been coded into a specific framework, like how math problems are solved. There are univerally accepted formulas for how mathematics works.

Supreme Court Justices, reimagined by Bing AI as D&D characters - part 2

For creating D&D character sheets, Bing AI had this disclaimer: Sure, I can try to create a character sheet for (name) based on his real-life attributes and achievements. However, this is not an official or accurate representation of him in AD&D terms, but rather a creative and hypothetical exercise. Please do not take this too seriously or literally.

Q: What is the probability of success in the D&D campaign Palace of the Vampire Queen with all the supreme court justices at level 1?

Bing AI:

First, let me assume that the campaign is run according to the original 1976 edition of the module, which was designed for 6-10 characters of levels 1-3. The module consists of five levels of a dungeon filled with various monsters, traps, and treasures. The final level contains the vampire queen and her coffin.

Second, let me assume that the supreme court justices are using the basic rules of D&D from 1974, which were compatible with the module. The basic rules allow for four classes: fighter, magic-user, cleric, and thief. Each class has different abilities and limitations. For example, fighters are good at combat and can use any weapon and armor, but have no magic; magic-users can cast spells, but are weak in combat and can only use daggers; clerics can cast spells and turn undead, but can’t use edged weapons; and thieves can pick locks, find traps, and backstab enemies, but have low hit points and limited weapon and armor choices.

Third, let me assign each justice a class based on their D&D stats, alignment, class, and guild from the previous table. I will also assign them a race (human or elf) based on their appearance and personality. Fourth, let me roll their hit points using the standard dice for each class: d8 for fighters and paladins, d6 for bards, clerics, druids, monks, and rogues, and d4 for wizards. I will also add their Constitution modifier to their hit points. 

Supreme Court Justices, reimagined by Bing AI as D&D characters - part 1

(obligatory attribution text) This content was generated by Microsoft Bing search chat mode, a conversational interface that can answer questions, generate creative content, and provide suggestions for the next user turn. To learn more about Bing chat mode, visit https://www.bing.com/chat.

Here is a possible table of D&D stats, alignment, class, and guild for the current 9 justices:

Name	Stats	Alignment	Class	Guild
John G. Roberts	STR 12 (+1), DEX 14 (+2), CON 14 (+2), INT 18 (+4), WIS 16 (+3), CHA 16 (+3)	Lawful Neutral	Bard	Order of the Harpers
Clarence Thomas	STR 12 (+1), DEX 18 (+4), CON 16 (+3), INT 16 (+3), WIS 20 (+5), CHA 12 (+1)	Lawful Neutral	Monk	Way of Tranquility
Samuel A. Alito	STR 14 (+2), DEX 10 (0), CON 16 (+3), INT 16 (+3), WIS 14 (+2), CHA 14 (+2)	Lawful Evil	Fighter	Battlemaster
Sonia Sotomayor	STR 10 (0), DEX 12 (+1), CON 14 (+2), INT 18 (+4), WIS 18 (+4), CHA 16 (+3)	Neutral Good	Wizard	School of Divination
Elena Kagan	STR 10 (0), DEX 14 (+2), CON 14 (+2), INT 20 (+5), WIS 16 (+3), CHA 18 (+4)	Neutral Good	Bard	College of Lore

Brett M. Kavanaugh	STR 12 (+1), DEX 10 (0), CON 14 (+2), INT 16 (+3), WIS 12 (+1), CHA 14 (+2)	Lawful Evil	Rogue	Assassin
Amy Coney Barrett	STR 10 (0), DEX 14 (+2), CON 12 (+1), INT 18 (+4), WIS 16 (+3), CHA 16 (+3)	Lawful Good	Cleric	Order Domain
Ketanji Brown Jackson	STR 10 (0), DEX 16 (+3), CON 14 (+2), INT 18 (+4), WIS 18 (+4), CHA 16 (+3)	Neutral Good	Druid	Circle of the Land