If the unit-economics work out and they can sell $0.99 of tokens for $1.00, doesn't matter how many agents you spin up. The flat rate subscriptions can't last though.
They are a luxury item, you are paying for the privilege of signaling you can afford $550 headphones. Generic black over-ear headphones could be $800, could be $80, useless for signaling. Doubly true in the context of a gift.
Ads are a ratchet that only tighten in one direction. Once the paychecks of 1000s of motivated, intelligent OpenAI employees depend on ad revenue increasing, the only option is to make them more invasive, more prevalent, more annoying, more data hungry etc.
That's not fair! Sometimes the ideas come from Snow Crash, which gave us the Metaverse because Zuckerberg wanted to cut a guy in half with a katana from a motorcycle.
Start with your chosen dictionary, remove small words and all 's'es.
Then find all words with 7 unique letters, then get the unique set of letter sets. These are your valid puzzles times 7 for each selection of a center letter.
Construct a trie of your dictionary. For each letter of each puzzle, walk the trie using only the puzzle letters. When you find a word, if it used the center letter, add to list.
The bigness of your data has always depended on the what you are doing with it.
Consider the following table of medical surgeries: date,physician_name, surgery_name,success.
"What are the top 10 most common surgeries?" - easy in bash
"Who are the top physicians (% success) in the last year for those surgeries?" - still easy in bash
"Which surgeries are most affected by physician experience?" - very hard in bash, requires calculating for every surgery how many times that physician had performed that surgery on that day, then compare low and high experience outcomes.
A researcher might see a smooth continuum of increasingly complex questions, but there are huge jumps in computational complexity. At 50gb dataset might be 'bigger' than a 2tb one if you are asking tough questions.
It's easier for a business to say "we use Spark for data processing", than "we build bespoke processing engines on a case by case basis".
Wish we lived in the universe where the term 'monkey' won over 'agent'. Would have given everything a cool Planet of the Apes feel.
I remember this getting a lot of buzz at the time, but few orgs are at the level of sophistication to implement chaos testing effectively.
Companies all want a robust DR strategy, but most outages are self-inflicted and time spent on DR would be better spent improving DX, testing, deployment and rollback.
The very next ask will be "order the zipcodes by number of customers" at which point you'll be back to aggregations, which is where you should have started