Hoang D.

Upper Moon at

[Hoang D. @pnjsilver]

@_chenglou charged testers we are, all of “them” change the games

[Hoang D. @pnjsilver]

@HarshithLucky3 may soon come to plus

[Hoang D. @pnjsilver]

@thsottiaux I think it's time to think about saving tokens.

[Hoang D. @pnjsilver]

@rezoundous when open, it does few queries, update interface, thats cost a lot tokens 🥷

[Hoang D. @pnjsilver]

@0xgaut jelly jelly

[Hoang D. @pnjsilver]

"CRITICAL INSTRUCTION 1: Always prioritize using the most specific tool you can for the task at hand. (a) NEVER run cat inside a bash command to create a new file or append to an existing file. (b) ALWAYS use grep_search instead of running grep inside a bash command unless absolutely needed. (c) DO NOT use ls for listing, cat for viewing, grep for finding, sed for replacing. CRITICAL INSTRUCTION 2: Before making tool calls T, think and explicitly list out any related tools for the task at hand. You can only execute a set of tools T if all other tools in the list are either more generic or cannot be used for the task at hand. ALWAYS START your thought with recalling critical instructions 1 and 2."

#Antigravity default instructions?

[Hoang D. @pnjsilver]

@_mohansolo mine not reset too,
sometime refreshes show in days, sometimes show in 5-6 hours, anyone has the same problem?

[Hoang D. @pnjsilver]

@Im_IrushiK an anchor to rely on to find another anchor, right or wrong according to the model's integrity, without responsibility.

[Hoang D. @pnjsilver]

It seems Generative AI is currently best suited for labeling.

What if AI evolution at this point hasn't actually gone beyond being just a sophisticated labeling machine?

[Hoang D. @pnjsilver]

🧐Premise:
7B smartphone users, 11B Facebook downloads.
1 user has m cycles of 1 buy + 1 sell + n other user re-sells

💡Formula: 7 * 10^9 * [( 1 * buy + 1 * sell + n * re-sell) * m] = 11 * 10^9

✅Simplified: (2 + n) * m = 11/7 ≈ 1.57

➡️3 Fun Estimation Scenarios for n & m:

Scenario 1: Everyone participates (m = 1)
- Result: n = -0.43
- Reality: Negative number. Average users only download to use (Buy). Very few actually market things (Sell), making global Re-sell behavior mathematically impossible in this model.

Scenario 2: Casual sellers (m = 0.5)
- Result: n ≈ 1.14
- Reality: Half the world is passive (download once & chill). The active 50% are online merchants who Buy + Sell, then flip or dropship (Re-sell) just over 1 time.

Scenario 3: The MMO & Ads Gods (m = 0.05)
- Result: n ≈ 29.4
- Reality: 95% of the world downloads FB exactly once. The massive 11B total is heavily carried by the top 5% of elite media buyers resetting devices and Re-selling (flipping via, ad assets) nearly 30 times! 🚀

[Hoang D. @pnjsilver]

@kritikakodes So, decades ago, we tried to convert words into code, and now we do the reverse.

[Hoang D. @pnjsilver]

@thenowhereway hi

[Hoang D. @pnjsilver]

@OfficialLoganK @roboflow I think, two biggest problems now are the calculating capacity and tokens limit rate, #ClaudeCode, #Codex, #Antigravity all the same. Models are not that top important.

[Hoang D. @pnjsilver]

Weekend and ppl not go out relax

[Hoang D. @pnjsilver]

Antigravity limit rate is funny, 8 messages and all gone #antigravity

[Hoang D. @pnjsilver]

@KaiXCreator give me the opinions and no code until i tell you to do

[Hoang D. @pnjsilver]

@iMessaage_ No way, how could the Santa climb down?

[Hoang D. @pnjsilver]

Just how bad my translation of a book, "A study in Scarlet". Should we have an evaluated translation app?
#translation #sherlockholmes #dichthuat

[Hoang D. @pnjsilver]

@zackdfilms1 good point to build laputa

[Hoang D. @pnjsilver]

@thsottiaux indeed a good plan, but coders now care much about limit rate and saving tokens