How to Use ChatGPT to Find Hidden Crypto Gems

Crypto Robot

What if ChatGPT could help you find the next 100x crypto before anyone else? 🚀 The combination of AI + crypto investing is one of the most powerful strategies for spotting hidden gems in today’s market — and it’s easier than you think.  In this video, I’ll show you exactly how to use ChatGPT to find hidden gems in the crypto market. From researching low-cap altcoins to analyzing narratives like DeFi, AI, and tokenization, ChatGPT can help you identify projects before they explode. We’ll also cover strategies for filtering out hype, using data-driven prompts, and combining ChatGPT research with tools like CoinGecko, CoinMarketCap, and TradingView.  Will ChatGPT give retail traders an edge in finding the next Bitcoin, Solana, or XRP before the crowd? Stick around as I break it down step by step and share how you can use AI to gain an edge in crypto investing.  👉 Do you think AI will play a huge role in crypto investing? Comment below with your thoughts — I’d love to hear them!  👉 Subscribe for daily alpha on crypto market trends, bold Bitcoin predictions, and altcoin gems that could 10x your portfolio! – https://www.youtube.com/channel/UCpjN8bNE-CoAgpfMatghM9g  📧 Email: cryptorobothelp@gmail.com  💰 Affiliate Links  Sofi Checking & Savings – Get $25 free ➝ https://www.sofi.com/invite/money?gcp=16a53d0f-b4b2-441d-9100-cfb506305260&isAliasGcp=false  Sofi Investing – Free $25 in stock ➝ https://www.sofi.com/invite/invest?gcp=ab31edd8-701e-4109-9225-51b41e35d246&isAliasGcp=false  Coinbase Exchange – Earn up to $300 BTC ➝ https://coinbase.com/join/YPUQLCY?src=referral-link  Tracking Tools – CoinGecko | CoinMarketCap  Trading Tools – Get $15 off TradingView ➝ https://www.tradingview.com/pricing/?share_your_love=cryptonextsteps  #ChatGPT #Crypto #Altcoins #CryptoNews #HiddenGems #CryptoInvesting #Blockchain #DeFi #Web3 #CryptoTraders #CryptoMarket #AI #CryptoUpdate #BullRun #CryptoStrategy

Transcript

00:00welcome to the deep dive where our goal is pretty simple yeah cut through the noise uh the

00:10information chaos and really hand you synthesized actionable knowledge to stay ahead that's the

00:17plan and today we are wading deep into crypto research i mean the market just generates noise

00:24faster than anyone could possibly process it really does and trying to find those genuinely

00:29overlooked opportunities those you know hidden gems that's often what separates the informed

00:33investor from well the rest of the crowd it's a classic dilemma isn't it due diligence is absolutely

00:39the foundation the bedrock of successful investing sure but in defy honestly it feels like a full-time

00:45job just tracking the data reading all the docs understanding the tech it's overwhelming it is so

00:50our mission for you today is really focused we're gonna define in detail a safe repeatable structured

00:56and maybe most importantly an ai accelerated workflow for finding these opportunities okay

01:01we want to turn that information overload problem into more of a processing problem something manageable

01:07ai accelerated okay let's untack this uh before we even bring in chat gpt or the llms we need to

01:13frame this to find the target when we say hidden gem for this deep dive what exactly are we hunting for

01:20yeah we need a really precise definition otherwise you know we're just chasing random pumps or

01:25speculation right so a hidden gem for this workflow it's got to be a small cap or maybe a mid cap token

01:32definitely far away from the big top 10 assets okay smaller coins smaller yes and crucially it must have

01:40specific verifiable positive attributes we're looking for clear evidence things like improving

01:47fundamentals real early traction or maybe overlooked catalyst technical or market ones things you can

01:52actually measure exactly verifiable and just as important it still needs to lack that high mainstream

01:57coverage if bloomberg or you know cnbc are talking about it daily too late yeah it's probably too late

02:03for the kind of edge we're looking for the lack of hype is almost as important as the strength of the

02:07fundamentals here so we're not just looking for cheap tokens that's not the goal we're looking for

02:12fundamentally sound projects that the market hasn't fully priced in yet probably because of

02:17lack of awareness or maybe just the info is hard to get exactly that we are hunting for alpha that

02:22genuinely requires research effort it's not just sitting there on the surface right and that brings

02:27us neatly to the ai connection why involve a tool like chat gpt or any large language model when crypto

02:34research really relies on real-time you know on-chain metrics isn't there a conflict yeah that's the

02:41challenge isn't it the data the on-chain stuff is real time but the documentation the white papers

02:48the audits the tokenomics docs they're static and often frankly massive and that's precisely where the

02:55llm's main value lies it's time compression specifically time compression in digesting complex text okay think

03:01about the hundreds of hours potentially saved by not having to manually scan a 50-page white paper

03:06just to find the vesting schedule or trying to compare governance mechanisms across three competing

03:11protocols by hand the ai can summarize those complex docs it can distill key risks generate research

03:18checklists based on the documentation and do rapid cross-project comparisons on criteria you define so

03:25the ai is basically handling the grunt work the reading the comparing allowing the human the researcher

03:32to focus purely on the analysis and the decision making part precisely it's an acceleration tool

03:37for your cognitive load however and this is absolutely the foundational caveat for this entire

03:42deep dive okay important point coming up yes ai must must be paired directly with verified real data

03:49sources if you skip the specialized data tools the massaris the dunes and just ask the llm for the current

03:55market cap or active users what happens you risk relying on outdated training data or worse what we call

04:00hallucinated facts things it just makes up because it doesn't have the live info i write fabricated

04:05information exactly the llm is only useful for synthesis when the inputs the numbers the links

04:12the metrics are collected from external verified often real-time sources like massari dune analytics nansen

04:19etc okay let's get into that ai edge versus its limitations then where does it really provide

04:25a genuine measurable advantage well if we isolate it the edge is strictly within the processing the

04:32structuring and the synthesis stage it absolutely cannot collect the primary data but boy can it

04:39organize complexity right so the best fits for llm acceleration are things like as we said deep

04:45document digestion can you give some more specific examples it's not just white papers you'd feed it

04:49right oh absolutely not think about security audits a typical smart contract audit report it might be

04:55dozens of pages long full of technical jargon dense findings yeah tough reads very tough you can feed

05:01that raw text to the lm and ask it specifically like ignoring the low severity stuff list the three

05:06highest risk potential vulnerabilities you found in this audit boom instant summary that's useful or

05:12think about the complex legal structure some protocols have the ai can help generate checklists to make

05:18sure you've covered all the compliance or regulatory risks mentioned in their legal docs and what about that

05:23explain like i'm five function i find that super helpful when i'm trying to wrap my head around some

05:28new complex technical thing it's incredibly valuable because it tests your own understanding right if you

05:35can't explain something simply you probably don't understand it well enough to put money into it good

05:39point the ai can take a really detailed explanation of i don't know a novel sharding architecture or some

05:46complex zero knowledge proof implementation and break it down into simple bullet points or analogies that acceleration of

05:53clarity it's often the difference between grasping a core tech advantage and just dismissing a project

05:58as too complicated yeah i can see that but let's be you know brutally honest about the limitations too

06:04this is the not a fit alone warning where do people commonly misuse these llms in crypto research

06:10okay the most common and honestly the most dangerous misuse is expecting real-time data or market

06:16sensitive insights directly from the llm like asking for the current price exactly chat gpt cannot give

06:23you real-time prices it cannot give you privileged data on wallet flows unless you feed it the output

06:30from a tool like nansen first because its training data is old by necessity yes it's delayed which means it

06:37absolutely must be used after the data is collected from those specialized real-time tools

06:43the massari screeners the dune dashboards the nansen label data these are non-negotiable prerequisites

06:50right relying only on the llm during the data gathering phase it's like trying to navigate a

06:54fighter jet using a map printed six months ago you're gonna crash you guarantee you'll crash into outdated or

07:00worse completely fabricated information that makes the relationship really clear then the ai speeds up

07:05the thinking the analysis the organization but it doesn't replace the initial verifiable data gathering

07:11it's like a world-class assistant for the researcher perfect analogy it takes the verified inputs you

07:16give it and structures them but it's not the detective out there collecting the evidence in the first

07:20place precisely if we use say a legal investigation analogy the llm is the expert paralegal organizes

07:29evidence cross-references documents drafts arguments okay but the dedicated data tools dune nansen

07:36massari they're the detectives collecting the fingerprints the surveillance footage in real time

07:42you absolutely must trust the detective work before you ask the paralegal to synthesize the case okay this

07:48clear division of labor brings us right to the core of this deep dive the repeatable structured five-step

07:54process this is the roadmap the workflow designed to efficiently move you from that raw overwhelming data

08:00to a concise risk-defined investment thesis let's start with the first three steps purely data gathering

08:05and shortlisting right step one source discovery let's call it the screening filter the goal here

08:10is just managing that info overload create a shortlist maybe 10 to 20 tokens max that fit our hidden gem

08:16profile small or mid-cap low coverage high potential exactly and the action here starting with neutral

08:23specialized data hubs screeners basically like the ones massari provides are a good starting point

08:29you're applying filters to quickly weed out projects that are either too large you know no alpha left or

08:35maybe too small too illiquid too risky you'd be surgical with those filters absolutely what are the specific

08:42like non-negotiable filters we should be using here okay first you need to define your sector don't try

08:48to boil the ocean focus on narratives the market is maybe just starting to track things like l2 infrastructure

08:54or deepen and decentralized physical infrastructure or rwa real world assets these are pretty focused

09:00areas right now but commit to one or two for a screening session focus is key okay what else second

09:06valuation check critically important you must filter using both market cap mc and fully diluted valuation

09:12fd why both because for small caps fdv is often the much better metric it accounts for all those future

09:18token unlocks the potential dilution down the road if the fdv is like astronomical compared to the

09:23current market cap even if the mc looks small danger sign huge danger sign the long-term sell pressure is

09:29likely toxic we want projects where the fdv isn't say 50 times the mc unless the tokenomics clearly show

09:38how value is captured proportionally to that dilution that fdv check is so critical so many people just see the

09:43small market cap and completely miss the massive dilution bomb waiting to go off yeah okay sector

09:49valuation yeah what else liquidity checks non-negotiable can you actually buy or sell a

09:55meaningful amount without tanking or pumping the price use the screeners liquidity metrics or check

10:00the order books on major exchanges if you can't get in or out reasonably you're trapped they're trapped

10:06and finally developer activity we need proof that the project is actually being built actively improved

10:12look at platforms like token terminal or sometimes github directly or other code tracking sites

10:17check the frequency of code commits the number of active core contributors over the last say 90 days

10:22so not just hype but actual work being done exactly if dev activity is flatlining the project is likely

10:28dying regardless of what twitter says high consistent dev activity is a strong proxy for future potential

10:34okay great so step one gives us a filtered shortlist active devs sensible valuation small cap projects

10:41now we need to prove they're actually being used not just held by speculators hoping for a pump

10:47that takes us to step two on-chain validation right this is where we shift gears from market level data

10:53down to granular usage data and for this powerful dashboards like those on dune analytics are

10:58indispensable the goal is to cut through the hype validate the project's actual utility is this token

11:05actually needed for the network to function is the fee revenue growing is anyone actually using this

11:10thing makes sense which specific panels or metrics on dune should we be prioritizing where do we focus

11:15you need to pull usage and holder data and critically track the trends not just a single snapshot in time

11:21okay trends over time yes the key metrics to pull maybe save links to the charts for are active

11:28addresses daily or weekly depending on the project's life cycle transaction count shows the actual network

11:34activity load new wallets created this can be a measure of expansion velocity and new user growth

11:40and potentially exchange flows are tokens moving on to exchanges to be sold or off exchanges into wallets

11:46for holding why do these specific metrics active addresses transaction count new wallets matter so much

11:53more than just looking at the price chart because they represent the verifiable difference between a

11:57temporary fad and an actual business being built think of it like this yeah imagine a new app or an l2

12:04or a game if its token price is pumping like crazy but the number of daily active addresses is totally

12:10flat it's like a trendy new coffee shop where everyone's outside taking instagram photos of the

12:14sign but nobody's actually buying coffee exactly the price the hype is high but the daily active

12:20transactions the coffee purchases are low it's likely a purely speculative play a fad not sustainable but

12:27conversely if the prices may be flat or even drifting down a bit yet active addresses are

12:33consistently climbing week after week that's the signal that's fundamental adoption that's a

12:39potentially healthy business being built under the radar it's latent growth just waiting for a

12:44narrative spark or wider market recognition okay and you mentioned tracking this over time

12:49absolutely paramount these metrics must be tracked over a sustained period we think four to 12 weeks

12:55is a reasonable minimum baseline for this kind of analysis right and you need to overlay notable

13:01news dates right onto those charts if active addresses suddenly spike for three days right after a big

13:07partnership announcement probably temporary hype exactly but if they rise consistently steadily across two

13:13months regardless of specific news events that's a fundamental trend you can potentially build a thesis on

13:19the long-term trend data is what separates durable utility from fleeting speculation okay fantastic we

13:26filtered the universe step one we've proven the project is active and being used step two but for

13:31real conviction especially in crypto we often want to know what the biggest smartest players are doing

13:37that brings us to the final data gathering step step three this smart money sanity check yes the goal

13:45here is straightforward verify accumulation or distribution patterns by players who are

13:49generally considered sophisticated they might have better research deeper network access maybe earlier

13:55information and how do we do that for this we rely heavily on specialized wallet labeling tools nansen is

14:01the best known example here we're specifically looking to see if wallets labeled as sophisticated entities

14:07think institutional funds known market makers prominent ecosystem investors or builders have been

14:12consistently accumulating the token or if they've been selling off exiting their positions okay but i need

14:18to push back here a little isn't there an inherent bias if we're waiting for nansen to label wallets

14:25and then identify sustained accumulation haven't we already missed the earliest cheapest entry point

14:33aren't we just following the leaders after they've already made their move that is a really critical

14:38question and the answer is nuanced yes you will almost certainly miss the very earliest ground floor

14:43entry point that's often reserved for vcs or seed investors who got in months or years before the token

14:49even exists publicly right our target here isn't necessarily being the absolute earliest investor our target is

14:55the investor who gets in before broad market coverage starts before the hype cycle really kicks in okay

15:00seeing smart money accumulation does two crucial things for us at this stage

15:04one their likely deeper due diligence serves as a confirmation signal for the positive fundamentals

15:09you spotted in step two it validates your own work right no sanity check exactly and two

15:15their presence their buying often helps stabilize liquidity and implicitly validates the longer

15:20term investment case it adds a layer of institutional credibility if you will so we're essentially

15:25trading a potential for the absolute first 10x gain for maybe a higher probability of a safer 5x gain

15:32it's a risk management trade-off that's a perfect way to put it it's about increasing conviction and

15:37reducing risk by confirming your thesis with the actions of sophisticated players but another big

15:44but okay you have to be highly critical when interpreting nansen data or any wallet tracking data

15:49don't just look at a single massive inflow transaction why not because that could be anything

15:55a fund rebalancing between wallets a treasury transfer for operations a large over-the-counter

16:00o tc deal that doesn't reflect market sentiment it's noisy so what should we focus on you need

16:05to focus on sustained repeat buys from multiple labeled wallets over that same 4 12-week period we

16:12discussed look for the pattern not the single snapshot are several known funds or market makers

16:18consistently adding to their positions week after week that shows real belief it shows sustained belief

16:23in the project's long-term value not just a reaction to some short-term hype this consistent

16:29accumulation pattern acts as our final filter before we unleash the ai for synthesis all right

16:35this is the turning point we've done the hard yards gathering verifiable data the screening step one

16:40the usage validation step two the smart money check step three now we transition to analysis and synthesis

16:47this is where the ai like chat gpt really starts to shine structuring and stress testing all that raw

16:53data we collected exactly we're moving into step four chat gpt synthesis and this is where the magic

16:59can happen but only because we put in the work to gather high quality ingredients first garbage in

17:03garbage out especially with ai right so the action here is feeding the lm your actual verified findings you

17:11literally paste in the key quantitative metrics active addresses up 35 over the last six weeks you give

17:17the links to the specific dune dashboards you use you paste critical excerpts from the tokenomics documents you

17:22found on masari or the project site this is that crucial distinction again we are not asking the llm

17:28to go find data in the wild we are asking it to process and structure the verified data we meticulously

17:35collected in steps one to three precisely we are leveraging its incredible organizational and language

17:41processing capabilities not its uh often flawed world knowledge base okay so we feed it the good stuff

17:47what do we ask it to do then you issue a comprehensive prompt you ask the llm to

17:51synthesize all this disparate data to generate four specific mandatory outputs these are designed

17:57to accelerate your own due diligence process significantly four outputs what are they first

18:02ask for a synthesized prioritized list of risks associated with the project these risks should be

18:07derived directly from the documentation you fed it like tokenomics flaws audit findings and potentially

18:13implied by the on-chain metrics like maybe slowing user growth despite price action okay risk list first

18:18makes sense second a detailed competitor map and comparison table you instruct it based on the technical

18:25docs and market positioning i provided create a feature-by-feature comparison table against its main competitors

18:31this forces the ai to structure the competitive landscape giving you instant context on relative strengths

18:37and weaknesses super useful for positioning third third emote analysis ask it based on the tech docs and market data

18:46what makes this project truly unique defensible or difficult for a well-funded competitor to replicate

18:51quickly is it network effects unique tech community what identifying the sustainable advantage and the

18:57fourth one and fourth this one is critically important falsification tests ask the ai given my bull case

19:04thesis which you state clearly define clear metric based tests that would definitively invalidate this bull case

19:10falsification test let's pause on that yeah that sounds like structured intellectual humility forcing

19:15yourself to define how you could be wrong can you give a concrete example like what would the prompt

19:19and the output look like sure let's say you're analyzing a new layer 2 solution your bull case

19:24hypothesis might be something like i believe project x will capture 10 market share of unique daily l2

19:31users within 12 months because its novel execution environment offers 50 lower gas fees than competitors

19:37a and b okay clear hypothesis you feed the ai this hypothesis along with the competitor data from

19:42the table it generated the user growth trends etc and you ask for falsification tests the ai might

19:49come back with something like falsification tests hashtag one if project x's total value locked tvl growth stalls or

19:56declines for more than 45 consecutive days while its main competitors a and b collectively capture more than

20:0250 percent of all new unique daily senders across the l2 sector during that same period the bull case based on fee advantage

20:08capturing market share is likely invalidated wow okay that is really specific and actionable it takes

20:14the decision making away from just vague feelings or hope and ties it directly to measurable observable

20:20metrics it tells you exactly when you need to admit you might be wrong which is often the hardest but

20:25most important thing for any investor to do and that structure that predefined i'm wrong if x happens

20:31leads us perfectly into the final step of the workflow step five risk controls and pre-mortem

20:38this is where we solidify the thesis with proactive defense and psychological preparation right this

20:43sounds like integrating behavioral finance directly into the analysis process using the ai for that

20:47exactly right you use the llm here almost purely as a structured thinking partner a sounding board you

20:53task it based on the risks that identified for you in step four to help you draft specific position sizing

20:58rules for example given the high unlock risk in month six limit initial position to one percent of

21:04portfolio okay you use it to help define specific non-emotional exit criteria maybe based on those

21:10falsification tests i will exit the position entirely if active addresses fall below 80 percent of their 90

21:15day peak and stay there for two consecutive weeks write it down commit to it and the pre-mortem exercise

21:22that sounds powerful but also maybe a bit counterintuitive yeah how do we ensure that running a pre-mortem

21:29imagining failure doesn't just subconsciously anchor us to that failure scenario and make us too scared

21:35to actually make the investment that's a really profound point actually the framing is key the

21:40pre-mortem must be framed as a forensic exercise not as some kind of self-fulfilling prophecy okay

21:46forensic exercise how the prompt shouldn't be will this fail it should be something like assume this

21:52investment has definitively failed six months from now based on the risks we identified in step four

21:57tokenomics competition adoption walk me through the most plausible sequence of events the narrative

22:04that led to that failure ah okay working backwards from assumed failure exactly by generating this

22:09logical failure narrative before you invest you achieve a couple of things you directly combat

22:14confirmation bias that very human tendency to only seek out data that supports your decision after

22:20you've made it right we all do that we do and when you're forced to logically construct how failure

22:25could happen using the risks you already identified you're sort of psychologically inoculating yourself

22:31against the emotional pain of loss if it does happen you've already lived it mentally it helps prevent

22:36clinging to a losing trade based on hope that is really powerful psychological risk management confronting

22:42and structuring the failure scenario before committing capital makes you prepared for both outcomes

22:47so to get this high quality analysis reliably out of the ai you can't just ask simple questions you

22:53need slightly more sophisticated prompts we've found three specific prompt engineering techniques or

22:59patterns that consistently yield more structured actionable analysis okay give us the patterns how do we

23:05elevate the quality of the l1 output from basic summary to real analysis okay first is the summarizer

23:11pattern but with a twist we use this for document digestion but add a critical edge the prompt

23:17looks something like this summarize this project's tokenomics its emissions schedule and particularly

23:23the unlock risks using these pasted documents or links provide them then based specifically on the

23:29potential supply shock from unlocks and the vesting periods you found list five specific red flags or

23:35events that could cause a major price drawdown ah so it's not just summarizing what is it it's summarize and

23:41then use that information to immediately identify potential problems related to supply exactly it forces the

23:46llm to move from passive digestion to active critical analysis focused on a key risk area supply shocks

23:53it turns a huge block of text into a prioritized bulleted warning system very efficient yeah that's

23:58efficiency right there going from what is it straight to what are the biggest landmines related to supply

24:04okay pattern number two second we call it the comparator pattern this is fantastic for getting quick

24:09context on unit economics and the competitive landscape the prompt needs to be highly structured something like

24:16create a detailed feature by feature table comparing project a our target versus its main competitors

24:22project b and project c include columns for consensus mechanism core fee model estimated treasury runway in

24:29months top holder concentration percentage and stated roadmap maturity for q4 use only the pasted notes and

24:36official links i am providing now okay very specific columns and restricting it to the provided info yes that

24:41prevents hallucination this instantly visualizes the competitive field it helps you understand relative

24:47value without spending hours manually building a comparison spreadsheet yourself you know if project

24:52a has maybe a three-year treasury runway based on its docs but competitor project b only has six months left

24:58that's a huge comparative insight right there instantly surfaced super valuable for relative assessment

25:03definitely superior to manual spreadsheet building for every single competitor okay and the third and final

25:08final pattern this one we call the signal to noise pattern or maybe the devil's advocate pattern

25:12it's specifically designed to stress test your own inherent human bias towards the project you've just spent

25:18hours researching ah fighting confirmation bias again how does this one work the prompt goes something

25:24like this okay here are the key on chain metrics i've verified eg active addresses up 20 percent month over

25:31month but unique senders are flat and here are the nansen wallet label observations eg smart money accumulation

25:38is minimal mostly retail buying now based only on this data argue against investing in this project right now

25:46specifically what negative narrative or interpretation would invalidate the apparent bull case even if the

25:51price is currently rising wow okay so you're explicitly telling the ai to build the bear case using the data you found

25:58even if some of it looks positive on the surface precisely it forces the ai and more importantly

26:03forces you to rigorously challenge your own bias to give significant weight to any contradictory evidence

26:08or alternative interpretations it ensure you're not just seeing what you want to see okay these patterns

26:13are great let's try to bring this entire five-step workflow including the ai synthesis to life can we walk

26:19through two hypothetical case studies maybe one success and one near miss where the process saves someone

26:25from a bad investment absolutely these kinds of stress tests really show the workflow in action all

26:31right case study one let's call it smart money confirms the thesis this is our hypothetical success

26:35story we start at step one the massari screener it flags let's say a layer two focused on data

26:41infrastructure ticks the boxes low fully diluted valuation relative to the actual fees it's generating

26:48and crucially it seems to be flying completely under the radar barely any mentions on social media

26:53low analyst coverage okay good candidate we move to step two on chain validation we pull up the

26:58relevant dune dashboards and the data is compelling we see a consistent undeniable upward trend daily

27:05active users and the total fees generated by the protocol have been rising steadily week over week

27:10for maybe nine consecutive weeks on nine weeks straight yeah a genuine sustained sign of adoption

27:16it strongly suggests the product market fit is real not just temporary hype people are actually using it and

27:21paying for it okay strong fundamental signal now for step three the nansen confirmation does smart money agree

27:29and let's say in this case step three seals the deal we check the nansen wallet labels for this token

27:35and we find evidence of repeat maybe small but sustained accumulation buys specifically from say

27:41five different named sophisticated venture funds or market makers happening over that same nine week period

27:47so not just one big whale buy but coordinated consistent accumulation from known smart players

27:53exactly that pattern provides significant validation so now we move to the ai synthesis steps four and five

27:59right you feed chat gpt the links to the dune dashboards showing the rising usage the summary of

28:04the nansen findings about fund accumulation and excerpts from the tokenomics explaining the fee capture

28:09the ai takes these inputs and what is it output it synthesizes them into a concise easily digestible investment

28:15thesis maybe highlighting the synergy between the growing user base and the validated smart money accumulation

28:23but crucially using maybe the summarizer pattern on the tokenomics docs the ai also spits out a clear

28:29prioritized risk list perhaps it flags a complex vesting schedule with a moderate unlock coming in four months

28:36this allows the investor to size their position appropriately acknowledging that specific risk alongside the clear growth signals

28:43the whole process which might have taken weeks of manual digging and analysis gets compressed potentially into a single focused afternoon

28:51using this workflow that's the dream scenario efficiency plus conviction now let's flip it case study to the near miss this is how the

28:58workflow saves you capital helps you identify and consciously avoid a potential trap let's say we find a token that's

29:04generating absolutely massive sustained hype on social media twitter's buzzing influencers are shilling it the price chart looks like a rocket ship the

29:11fomo is intense ah yes the classic hype cycle okay but we stick religiously to the workflow no exceptions

29:19we move to step two the on-chain validation check and uh oh we find a stark contradiction between the social narrative

29:27and the actual usage reality what is dune show dune analytics shows may be flat or even slightly declining

29:34active addresses over the past month and perhaps more worryingly a sharp drop in the number of unique

29:39senders fewer distinct wallets are actually interacting with the protocol so lots of talk but usage is

29:45stagnant or falling big disconnect huge disconnect people are talking but they aren't actually using

29:50the product then we check step three nansen what's the smart money doing amidst this hype and nansen confirms

29:57the worrying usage trend it shows clear net outflows from wallets labeled as smart money over the past month

30:02they're selling into the hype not buying wow so the verifiable data completely contradicts the social media

30:07narrative the noise is deafening but the actual signal is flashing bright red how does the ai help

30:13us navigate this in the risk phase steps four and five okay the user seeing these conflicting signals hype

30:19versus reality retail fomo versus smart money selling decides to run a heavy risk mitigation step using the ai

30:27they specifically ask chat gpt step five to help draft a pre-mortem analysis okay and as input for this

30:33pre-mortem they feed the ai the tokens detailed vesting schedule and supply information which

30:39let's say was incredibly complex and buried deep in some governance forum post not easily visible getting

30:45complexity exactly the ai perhaps using that summarizer pattern we discussed digests this complex supply

30:51info and immediately spots a major issue a huge previously unnoticed emissions overhang

30:57a massive multi-million dollar chunk of tokens is scheduled to unlock and hit the market next month

31:03the unlock bomb the unlock bomb completely obscured by all the social media hype the ai then helps

31:08outline the highly probable failure scenario the token price likely tanks hard upon this massive unlock

31:14regardless of fundamentals which were already questionable based on usage data and the result the result

31:18is clear the user armed with this structured analysis of the usage data smart money flows and the hidden

31:24supply risk surfaced by the ai confidently passes on the investment they avoid a potentially massive

31:31drawdown triggered by that mechanical supply shock that the hype mob completely missed that right there

31:36is the ultimate value proposition isn't it the workflow amplified by the ai didn't just tell you what

31:41to buy it accelerated and structured your decision to not buy which let's be honest is often the most

31:47profitable decision an investor can make in crypto saving capital is key absolutely and to execute

31:54this kind of advanced ai assisted strategy successfully week after week you need to be

31:59really surgical and rigorous about the specific metrics you track consistently we rely on what we

32:04call three mandatory data panels these move way beyond just simple price and volume okay let's break

32:09down these panels panel one the on-chain traction panel this is about proving real utility right

32:14correct here we prioritize those three key metrics we discussed active addresses transaction count and unique

32:20senders these are the vital signs of a protocol's health you know if a million addresses hold a token but only

32:28a few hundred are actually active daily it's mostly dormant holders not users exactly you've achieved

32:33distribution success maybe but not utility success and again we track this data over 4 12 weeks minimum

32:40daily volatility in these metrics is mostly noise the sustained multi-week trend is the signal we care about

32:47any more advanced metrics within this panel yeah for a more sophisticated look especially within layer two

32:52ecosystems you might dig deeper than just total transaction volume look at the distribution of gas usage among the different

32:58protocols built on that l2 is your target project consistently driving a significant maybe growing

33:04percentage of the total gas consumed on the network ah so is it becoming indispensable to the ecosystem precisely that

33:11suggests deep integration and real demand not just transient speculative traffic hopping between dApps okay

33:17great panel two panel two holder quality and concentration this is where the nansen layer or similar analytics comes

33:24in heavily it's about analyzing who owns the supply and critically where it's moving what are we tracking here

33:31first track the top 10 or top 20 holder percentage over time if just a handful of early whales or the team wallet

33:38dominate the liquid supply big red flag centralization risk huge red flag centralization risk potential exit

33:44liquidity issues down the line but more specifically we track the number of distinct wallets labeled as fun

33:50market maker or smart money that are net accumulating versus those that are net distributing each week or month

33:56tracking the count of smart wallets accumulating not just the total volume yes because it shows the breadth of

34:01conviction is it just one fun buying or are multiple sophisticated players seeing the same opportunity

34:08this allows us to construct something really powerful we call the smart money heat map smart money heat

34:13map tell me more that sounds advanced it is we track these weekly net flows from labeled smart money

34:18wallets but we segment them by protocol sector l2s dempin rwa gaming defi lending etc why track fund flows by

34:27sector what does that tell us it helps you spot potential sector rotations early often before the narrative hits

34:34mainstream crypto media or influencers for instance if you consistently see net outflows from wallets labeled as

34:41active and defi lending protocols for several weeks but at the same time you see sustained net inflows into

34:49wallets known to be active and depend in rwa projects that suggests the big money is repositioning shifting focus

34:56exactly it suggests where sophisticated capital believes the next major growth narrative might emerge

35:03this gives you valuable lead time helping you prioritize opportunities from your step one screened list that align with

35:08these nascent institutional flows that's a really powerful edge okay panel three panel three a mission and unlock calendar

35:15this is all about the unit economics of supply understanding dilution is non-negotiable if you want to avoid being blindsided

35:21where do we get this data reliably this is primarily derived from meticulous reading of official tokenomics documents

35:27governance forums and often aggregated by data providers like massari you need to map out the

35:32cumulative circulating supply versus the fully diluted valuation fdv at key future dates mapping out future

35:40supply yes and specifically you must calculate the expected cell pressure from major upcoming scheduled unlocks

35:47for example if your analysis reveals that say 50 of the total remaining unvested token supply

35:53unlocks for early investors in the team over the next four months you need to quantify that you need

35:58to quantify the approximate usd value of that unlock at current prices to understand the sheer scale of

36:03the potential mechanical cell pressure hitting the market the ai can help process the complex unlock

36:09scheduled details from the docs but the human researcher has to decide if the fundamental growth story from step two

36:15is plausibly strong enough to absorb that incoming supply without cratering the price right growth versus

36:20supply pressure it's a balancing act yeah now this whole workflow is designed to find hidden gems which

36:27often means smaller market caps potentially lower liquidity we absolutely must talk about the safety

36:33checks and ethical considerations here could agree more the final safety requirements are paramount especially

36:39in this segment of the market first you must be hyper alert for potential pump and dump dynamics these thrive

36:46in low liquidity environments a quick check can a relatively small investment small relative to your

36:52typical position size significantly move the price of the token on the main exchanges if yes if yes that

36:58signals a thin liquidity environment it's a major risk your investment itself could be potentially illiquid when

37:03you need to exit proceed with extreme caution or more likely just pass makes sense and what about that

37:09concentration rule you mentioned derived from panel 2 data yes the holder concentration if your analysis

37:14using nansen or similar tools reveals that the top say 10 non-exchange non-protocol wallets control a

37:22significant overwhelming percentage of the liquid supply the definition of significant depends on your risk

37:28tolerance but generally anything above 50 concentrated in so few hands is highly dangerous

37:33what's the verdict it's a clear unambiguous risk indicator the potential for manipulation for coordinated

37:39dumping or just massive cell pressure if one whale exits is simply too high our definitive recommendation

37:45based on analyzing countless past failures is usually just to pass on that opportunity find something with

37:50healthier distribution and the final guiding principle here the mantra that must be followed

37:55religiously throughout this entire process use ai to explain risk never ever to ignore it the llm should

38:02make you more cautious more structured more aware of potential downsides not less disciplined or more

38:08prone to fomo so just to recap our journey today the real transformative power of ai in crypto research as we

38:15see it isn't really in primary data collection it's purely in the synthesis the structured analysis and maybe most

38:22of the results uniquely in facilitating that psychological risk management piece right it takes that complex verified data you feed it

38:29and helps rapidly construct a clear defensible thesis and just as importantly a clear risk profile

38:36but the power of the overall workflow itself lies in its discipline structure combining the high-level screening from tools like

38:42massari with the granular usage proof from dune and that critical smart money validation from nansen to work together they have to work together

38:49without those verifiable data inputs feeding the machine the ai is at best a sophisticated distraction and at

38:55worst a dangerous source of confident sounding misinformation the key takeaway then is really

39:00understanding that these tools the data platforms and the ai are separate but they're critically interdependent you need

39:07the whole chain when used correctly together they form a pretty robust shield against market noise

39:13hype and purely emotional trading decisions well said so as we wrap up we want to leave you with a

39:18provocative thought something to mull over forcing you to choose signal over noise in your own process

39:25imagine this scenario nansen's labeled smart money wallets are clearly actively exiting a token you're

39:31watching net outflows week after week but at the same time social media sentiment is absolutely

39:37mooning it's pure rocket emojis to the moon posts fueled by hype and fomo which signal do you follow

39:42the verifiable flow of sophisticated capital hitting for the exits or the deafening roar of the crowd

39:48narrative that's the kind of hard question this structured workflow forces you to confront and

39:53hopefully answer rationally long before you click buy that's the test we really hope this deep dive into ai

39:59assisted crypto research has been valuable maybe giving you a concrete framework or refining your own

40:05process if this workflow helped you define your search criteria better or gave you specific metrics

40:11to start tracking or maybe just clarified how to actually use these powerful tools like chat gpt nansen

40:19and dune together yeah well we need your help too yeah absolutely engaging with this deep dive yeah and

40:26subscribing to the channel if you haven't already dropping a comment below maybe letting us know which metric

40:30you think is the absolute must have before buying or just hitting that light button it genuinely helps

40:36support us it really does it boosts our visibility in the algorithms lets youtube know you value this

40:41kind of content and frankly it allows us to keep creating this sort of quality research heavy crypto

40:46analysis for you we love doing it but your engagement makes it possible we appreciate it so thanks

40:51for diving deep with us today we'll catch you on the next one

41:00you

Category

Transcript

Be the first to comment

Recommended