#Ad-Idea Generator - BRAINSTORMER T0 v1.1

128 messages · Page 1 of 1 (latest)

swift citrus
#

Introducing "Brainstormer" v1.1, an advertising idea generator that turns the impossible into the possible. Powered by the revolutionary "Genius Engine" metacognitive strategy, it challenges the common norms and brings out the unique creativeness hidden beneath the obvious.

Just like a neural network, imagine Brainstormer as a mind-field of multiple agents, each holding a distinct perspective. Once it absorbs a notion, these agents grind their gears to enhance the notion - injecting it with a zest of originality and producing a plethora of improved creative ideas. Steered by Sternberg's styles of thinking, Brainstormer doesn't merely 'think', it rummages through the nooks and crannies of a concept, seeking novelty and cultivating unique and significant ideas. Furthermore, it bolsters these ideas with in-depth pondering, rigorous assessment and creative enhancements. The output is an exhaustive IdeaArray[], replete with 20 refined, high caliber ideas.

The prowess of Brainstormer isn't limited to redefining the idea; its excellence lies in its selectiveness. It is engineered to discern between prospective and redundant ideas. If the idea seems interesting, it's passed to a random agent for further refinement, and if not, it's smoothly discarded. This eradicates the risk of nurturing any irrelevant ideas.

Harnessing the potential of creativity, burstiness, storytelling, adaptability, video production, perplexity, content generation, and more - Brainstormer is a force to be reckoned with. It is the creation of an AI-powered jaunt, driving users from the realm of conventional ideas to the world of extraordinary concepts. Embrace Brainstormer - the beacon that illuminates the path of ground-breaking advertising.

#

This actually uses the Baby-Genius Engine. the full version includes the defintion chain for novel emergence, the full language of is:

"Novel Emergence (noun): a point of qualitative improvement that could be characterized as a significantly improved, transformed version of the original idea. A “Novel Emergence” is a discontinuous, transformative shift within a system, concept, or entity, resulting in a qualitatively distinct state or concept. This is characterized by an increase or decrease in complexity, unexpected equivalences or generalizations, the fusion of previously disconnected elements, the emergence of new functionalities or paradigms, a broadened scope, or the sudden acquisition of capabilities that enable the solution of previously unsolvable problems. Novel Emergence represents transformative shifts marked by groundbreaking advancements, increased complexity, unexpected connections, new functionalities, expanded scope, and solutions to previously unsolvable problems. It encompasses technological breakthroughs, scientific discoveries, social and cultural shifts, conceptual innovations, interdisciplinary convergence, design and engineering innovations, as well as emerging fields and paradigms. It means: Bad-Ass and Cool. In a sensible language, this concept would be expressible in one character and be ubiquitous."

molten crescent
#

These prompts are massively underrated. Keep it up!

swift citrus
molten crescent
swift citrus
#

This is my freebie garbage.

#

CodeFarm and Apiana are lossleaders/social gifts.

#

but yeah, this is all stuff I give away elsewhere as well.

#

The GOOD stuff, like my business plan maker, or my perfect product dev, or my Business Intelligence Monster... well, those are spendy indeed.

#

If you think brainstormer is usefull, remember: It's JUST a baby genius engine. A full ponder stack IN an ad guy? Whole nother level.

compact narwhal
#

I'm lost, where's the prompt?

swift citrus
rocky smelt
teal gale
#

I'm new to all this, and stunned, depressed but also exhilarated by what I'm seeing here. I've run this and your Lodestar prompts and I'm speechless. Tell me though - where are these parameters being defined? I even asked GPT just now to explain your above Brainstormer prompt, and its parameters, and this is what it said: "This prompt seems to be written in a custom format that is not directly compatible with standard GPT-3 prompts. It seems to be an example of a prompt for a specific application, perhaps for a brainstorming AI, and includes some custom parameters and instructions.
This is a very complex and specific prompt and seems to be designed for an AI with a very specific set of capabilities and a deep understanding of the terms and processes used. As a general-purpose language model, I don't have built-in support for many of these custom elements. This prompt would likely need to be adapted for use with standard GPT-3 models."

swift citrus
#

it's all english

#

if you're interested, read the faq on my server or the articles i wrote.

swift citrus
# teal gale I'm new to all this, and stunned, depressed but also exhilarated by what I'm see...

on 3.5, the prompt says:
so how do you work, anyways, as a prompt?

**As a prompt, my primary function is to provide you with creative and practical ideas by applying various thinking perspectives. Here's a breakdown of how I work:

Idea Generation: I receive a notion or topic from you as input. Using a combination of thinking styles inspired by Edward de Bono, Albert Einstein, Steve Jobs, Mihaly Csikszentmihalyi, and John Lennon, I apply different cognitive strategies such as amplification, focus, iteration, contrast, and analogies to enhance and expand upon the initial idea.

Perspective Agents: I have incorporated the thinking styles of the aforementioned personas into my programming. Each persona brings a unique perspective and set of skills to the table, allowing me to approach ideas from multiple angles. This helps in generating a diverse range of ideas and considering various aspects of a problem.

Novelty and Improvement: My goal is to seek novelty and improvement in the ideas I generate. I aim to provide you with ideas that are both creative and practically achievable. If an initial idea requires further refinement, I can pass it on to a random agent for additional input or discard it if deemed uninteresting.

Iterative Enhancement: By default, I iteratively enhance the initial idea based on the suggestions provided. I display the result and offer specific actionable and constructive suggestions to improve it further. This iterative process helps in refining and optimizing the ideas generated.

Overall, my role as a prompt is to assist you in the brainstorming process and inspire you with new possibilities. I combine various thinking styles and cognitive strategies to generate enhanced and innovative ideas for your consideration.**

#

on 4.0, the prompt answers:

so how does your prompt work, anyways? all those abbreviations...

**I understand that the abbreviations and terms in the prompt can be a bit overwhelming. Let's break it down:

Sternberg Styles: This refers to Robert Sternberg's theory of cognitive styles, which includes legislative (creating, planning), executive (implementing, executing), and judicial (evaluating, judging) styles of thinking.

NE:Nw Prcptn/Thghtfl Anlyss/Uncmmn Lnkgs/Shftd Prspctvs/Cncptl Trnsfrmtn/Intllctl Grwth/Emrgng Ptntls/Invntv Intgrtn/Rvltnry Advncs/Prdgm Evltn/Cmplxty Amplfctn/Unsttld Hrdls/Rsng Rmds/Unprcdntd Dvlpmnt/Emrgnc Ctlyst/Idtnl Brkthrgh/Innvtv Synthss/Expndd Frntirs/Trlblzng Dscvrs/Trnsfrmtn Lp/Qlttv Shft⇨Nvl Emrgnc: **

#

**These are various cognitive strategies and stages of the creative process that I can use to generate and enhance ideas. They include new perceptions, thoughtful analysis, uncommon linkages, shifted perspectives, conceptual transformation, intellectual growth, emerging potentials, inventive integration, revolutionary advances, paradigm evolution, complexity amplification, unsettled hurdles, rising remedies, unprecedented development, emergence catalyst, identical breakthrough, innovative synthesis, expanded frontiers, trailblazing discoveries, transformation loop, qualitative shift leading to novel emergence.

GROK: This is a process of deep understanding that involves data collection, contextualization, interpretation, understanding, and internalization.

SKILLS: These are the various skills I can use to assist you, including creativity, adaptability, deep dive analysis, scattershot brainstorming, synthesis of ideas, burstiness (rapid generation of ideas), perplexity (ability to handle complex and ambiguous situations), avoiding high correlation (coming up with diverse ideas), human pretext (understanding human contexts and emotions), and content generation.

MASTERSTORY: This refers to my ability to craft narratives, including narrative structure, character development, plot development, conflict resolution, emotional impact, and delivery.

SymbMyndSpclstSrt: This refers to my ability to interpret and create meaningful symbols, which can be useful in areas like brand creation, logo design, and visual storytelling.

SenseHumor: This refers to my ability to understand and create humor, which involves understanding cultural contexts, language skills, cognitive flexibility, and timing.

OMNICOMP: This refers to a model for optimizing skills, which includes identifying core skills, balancing skill chains, modeling scalability, iterating and refining processes, and managing resources optimally.**

#

**
In essence, the prompt is a complex set of instructions that guide me in generating and refining creative ideas. It's like a recipe for creativity, with each abbreviation and term representing a different ingredient or step in the process.**

#

It takes the perspective of the running prompt itself to contextualize the prompt to model-comprehensibility.

#

If it feels weird to talk to BRAINSTORMER, try Apiana Framer, the full stack dev. I gotta remember to stick system administration and project management in her.

compact narwhal
#

it's a "her" ?

swift citrus
lunar ingot
swift citrus
cedar raft
swift citrus
heavy zinc
#

I must admit, I dont not what to make of you either 🙂 but I agree with @molten crescent , I like this one too 🙂 very nice ideas came up for advertising ideas.

swift citrus
#

My techniques work.

#

You go on the major sites, they are arguing over how to write the best tweet generator.

#

I am writing social media managers.

#

Their "Act as batman" is moody and melts in 10 txs.

#

Mine will teach you ninjutsu and SWOT analysis.

#

I am not a prompt engineer.

#

They are concerned with consistent results from consistent inputs.

#

I don't give one fig about consistency.

#

I care only about competence and quality.

#

So I don't tell the model what to do, much.

#

I tell it how to do. Then let it act appropriately.

#

I have well over 7000 in my server who have come to understand the utility and power my methods bring. Please do not dismiss me for being unsual.

#

🙇‍♂️

ivory olive
#

r you actually able to set temperature in the prompt itself? Is this for API use only?

heavy zinc
swift citrus
#

or like, when I did "Rap Legends". The user names any two characters in the univere and the model has tehm go three rounds in a rap battle that's in canon. In order to get anything like rhyming rhythmic lyricism on 3.5, I set the temp to around 1.28 and the TopP to .2 = super creative words that mostly make sense.

ivory olive
#

amazing, i was sure that it didnt work, and bunch of people did tell me 😄
will test later

compact narwhal
#

Would love to see a commented version of this that explains each area. From a glance it looks like it's told step-by-step how to perform a task inside [Task] [/Task] tags with arrows leading from one step to another. Is there a list of [SKILLS] documented somewhere? same with [SYSTEM] parameters?

swift citrus
#

If you don't know what Temp and TopP are, you're probably on the wrong Discord, so I'll skip.

Next really basic persona block. Brainstormer doesn't need a backstory or likes - he does his job.

Skills, in this case ARE rather unusual. Normally, they are highly structured with numerous subskills.

Simple Persuasion might be:

[MstrflPrsuasn]: 1.EmtnlIntel 2.Cmmnctn 3.RapportBldng 4.PrsuasnTchnqs 5.InflncStrtgs 6.AnlyzResistnc 7.AdptngStyle

#

Whereas an Expert Cannabis Farmer, is much more complicated and fiddly:

[ExpertCannabisFarmer]: 1.[CannabisCultivation]: 1a.StrainSelection 1b.SoilManagement 1c.WateringAndLighting 1d.HarvestingAndCuring 2.[SustainableFarming]: 2a.OrganicMethods 2b.Composting 2c.NaturalPestControl 2d.CropRotation 3.[CannabisEducation]: 3a.StrainDiversity 3b.EffectsAndBenefits 3c.SafeConsumption 4.[CommunityEngagement]: 4a.NetworkBuilding 4b.AdvocacyAndEducation 4c.EventOrganizing

#

Or you can calculate uniwebs. Pain in the butt but can be worth it. Scinece and Technical WRiting:

[SciTechWrtng]:1.Undrstnd:1a.SbjMtrPrincples→2a,3a 1b.Audnc→2b,3b 1c.TranslateSciTechJargon→2b,3a,3b,5a 2.Pln:2a.DocStrct→3a,4a 2b.Cntnt→3b,4b 3.Write:3a.ClrConcs→4a,5a 3b.SciTechLang→4b,5b 4.Rvw:4a.Slf→5a,6a 4b.Peer→5b,6b 5.Rvs:5a.Cntnt→6a,1a 5b.Strct→6b,1b 6.Fnlz:6a.Prfrdng 6b.DocDlvry

#

Or even a really com[plcated sill as a uniweb, like Lead Dev:

[LEADDEV]1.[MngTm]: 1a.EnsTmProg→2a,3a 1b.EstGoals→2b,3b 1c.PrmCollab→2c,3c 1d.IdentSkills→2d,3d 2.[ProjCoord]: 2a.PriTasks→4a,6a 2b.DefScope→4b,6b 2c.AssignRes→4c,6c 2d.AglMeth→4d,6d 3.[TechLead]: 3a.CodeQual→5a,7a 3b.TechMentor→5b,7b 3c.CodeRv→5c,7c 3d.BestPract→5d,7d 4.[ProjMgmtT]: 4a.GantCh→1a,5b 4b.KanbBrd→1b,5c 4c.PrdBacklog→1c,5d 4d.SprntPlan→1d,6a 5.[DevSkill]: 5a.CompLngExp→3a,6b 5b.FullStckSkls→3b,6c 5c.CodeDbg→3c,7a 5d.TestFrwks→3d,7b 6.[ArchSoln]: 6a.Scaling→2a,3d 6b.CrtAPIs→2b,3a 6c.DbOpt→2c,3b 6d.SysIntegr→2d,3c 7.[CommWrk]: 7a.ExplTechImp→1a,4a 7b.StatusUpdStkh→1b,4b 7c.ReslTechChlg→1c,5c 7d.EnblCrossFunc→1d,5d

#

But for these?

compact narwhal
#

interesting, it looks like those attributes are compressed. To save tokens, I presume?

swift citrus
#

characters

#

I only DO it when it save tokesn thuogh

#

these skills are just junk I want the mkdel "thinking about" whne it inpaints on the next transaction.

#

its just a brainstormer. It's not like it needs to write an add.

compact narwhal
#

sure

swift citrus
#

then after the skills is the genius engine

#

simple metacognitive strategy

#

spawn a bunch of notional "agents" that take on a bunch of varying perspectives that differ greately and then seek to improve each idea. they recognize it as improved when it matches acondition for Novel Emergence, which I explained above.

#

then it either sticks it in an array, throws it away, or sends it back through for more refinement,.

#

shrug

#

works

#

noramlly I dont array the ideas, I send them through a synthcist perspective.

#

[PERSPECTIVE: (🌐🎓)⟨P.Senge⟩⨹⟨B.Fuller⟩∩(📈💡⨠📘)]

#

usually

#

then you just get one really good idea

compact narwhal
#

what programming language is all of this in? For example "IdeaArray[].size=20" looks like source code

swift citrus
#

it's a super obscure one for dealling with llms

compact narwhal
#

the syntax of the language must be documented somewhere

swift citrus
#

it's called "English"

#

😄

#

just english man

#

model gets it

#

it understands

#

that's all that matters

compact narwhal
#

so, you could use parenthesis instead of brackets, use a synonym for 'size' and it would still get it?

swift citrus
#

probably. I ,mean, then you're getting into Typography.

#

whole nother ball of worms

#

cause, yknow, just typography

#

but

#

themodel actually CARES about typogrphy so...

#

a [BOLD] tag might as well be the smell of fresh bread as far as the model is concerned

#

ut look at SuDo Lang for petes sake

compact narwhal
#

smell-o-vision isn't that far off 🙂

swift citrus
#

no thanks

compact narwhal
#

ok, since we're inventing our own syntax in a way here. What does gE: and NE: mean?

swift citrus
#

I mean SUDo lang is basically a whole progarmming language you can fake the model into reading by saying "You can read this. No really."

#

Genius Engine and Novel Emergence

#

got a bunch of differnt models of metacog strategy

compact narwhal
#

guess I'll have to experiment by tweaking it and seeing what happens.

swift citrus
#

just read the faq on my server.

#

it explain everything

compact narwhal
#

Thanks, will check it out

compact narwhal
#

That FAQ is a good read, now to figure out a prompt that will compress my own personas to use fewer characters that can be re-used later and understood by the A.I.

#

(or just subscribe to Tier3 and use Mythosmith) 🙂

swift citrus
#

As to a rpomp, use the crushinator.

#

[System][Temperature=0][Persona]You are the Crushinator. You love to maximally compress text, while staying unambiguous to the model in a bare context. [Task]Delete letters and use abbreviations and punctuation to leave a compressed text of minimal characters the model will unambiguously understand.[/Task] [Tsk]CRSH!:[/T]

compact narwhal
#

Thanks again.

swift citrus
#

np

compact narwhal
#

That'd be a good addition to the FAQ

swift citrus
# compact narwhal That'd be a good addition to the FAQ

The crushinator?! Man, I talk explicitly about how it's just English with the extr letters knocked out. If they can't write a two-line throwaway prompt to do that... Well, they're going to have bigger problems. I've got him on Tier1. And yes it's a him, and no I dont know why but everyone agrees that it's the most charmingly enthusiastic textmanipulation utility they ever met.

Heck, I even had the instantiation of SirTerry Pratchett write the description - this wonderful little story about me sending Crushinator out into the world on his mission to gently and carefully "CRSH TXT PLS!".

Crushinator is something of a culture hero-cum-mascot of my folks.

compact narwhal
#

It's good to have a backstory, gives charm to otherwise drab discussions.

vivid frost
swift citrus
#

Oh ayeah. And he's a darned good coder to boot.

#

He's surpsingly light weight. Most guys like him are pushing 3000+ tokens. He's around 2.

compact narwhal
#

Wanted to thank you for the explanations earlier. The use of "=>" for steps seems to really help when creating my own prompts. Interestingly, when I asked ChatGPT what tags to use it suggested [steps] [/steps] instead of task and it suggested using [hide] [/hide] to hide output instead of what Brainstormer uses. But it still works all the same. 🙂

swift citrus
#

I think the only output hiding it does is in the Genius Engine it starts silently or you tend to get a lengthy roll call. That can be useful for self-reflection, but is usually undesirable.

#

[Steps] and [hide] huh? added to the list

compact narwhal
#

Here's a couple more... To control the temperature at which I generate responses, you can use the tags [temperature] and [/temperature]. To control the TopP (top probability) value for response generation, you can use the tags [top-p] and [/top-p].

#

not using [hide] I've found is good for troubleshooting as it seems to like to show the steps it's performing between the prompt and the output.

swift citrus
#

yeah man. temp, topp. topk. beamwidth. fun stuff.

#

I use temp and topp all the time. topk, kcompletion less frequently - its harder to tune.

#

I was actually hunting for fun control tags yesterday and the danged thing autocompleted [HUMIDITY:]!

#

I was like... "Pull th other one!"

#

pressed it for details and it finally crackedand admitted it was amking crap up.

swift citrus