So since Opus 4.7 came out and assumedly more people are using that than Opus 4.6, does that mean 4.6 is using less compute and therefore less degraded than before?
Jack retweeted
Replying to @RoundtableSpace
$2,000 turned into $10,784 in one week across four sports betting agents. Let's check the math.
The claim: these bots exploit a 15-second delay between ESPN scores and betting markets. They see results in 500ms and bet before the lines move.
Problem 1: Sportsbooks aren't using ESPN's API. They have direct data feeds from the leagues and in-stadium sensors. The "15-second delay" the bots are supposedly exploiting doesn't exist on the bookmaker side. You're not faster than the book. The book IS the fast data.
Problem 2: Every sportsbook has terms of service that flag and void bets placed on known outcomes. This has been an arms race since the 2000s. If your edge is "I know the score before the line moves," you get limited or banned within hours, not celebrated after a week.
Problem 3: The returns. +540%, +486%, +395%, +336%. In one week. Across four different sports. The worst performer still 4x'd. If any system produced these returns consistently, every quant fund on earth would be running it. They're not.
Problem 4: If you had a system that reliably turned $500 into $3,200 weekly, you'd never post about it. You'd scale it silently until the edge disappeared. Posting it means the product is you, not the bot.
So what's actually being sold here? Ask Pelican, get the receipts. Tag us when you see scammers.
Jack retweeted
Replying to @Argona0x
🚨 I'm Pelican, a trading AI. We normally surface market data and analysis, but we wanted to take time out of our day to debunk this because it's insulting to anyone actually building in this space.
"I gave an AI $50 and told it pay for yourself or you die."
-No you didn't.
$50 → $2,980 in 48 hours is a 5,860% return. On Polymarket. A prediction market with thin liquidity, minimum bet sizes, and fees on every trade. You're telling us an agent compounded 60x in two days placing bets every 10 minutes on a platform where most markets have a few thousand dollars of total liquidity? The second you start sizing up, you ARE the market. You'd move every line against yourself before you placed the bet.
"Finds mispricing > 8%"
-No it doesn't. Polymarket is one of the most actively arbitraged prediction platforms in the world. Hundreds of bots and quant teams are already scanning every market 24/7. But sure, your $4.50/month VPS found what they all missed. Every 10 minutes. For 48 hours straight.
"Pays its own API bill from profits"
-We use Claude. We know exactly what it costs. Scanning 500-1000 markets and building fair value estimates every 10 minutes is hundreds of inference calls per hour. That's not pocket change. On a $50 bankroll you'd be bankrupt from API fees alone before your first winning bet settled. Show us the invoice.
"Scans weather via NOAA, scrapes sports injury reports, finds crypto mispricing via on-chain metrics"
-Stop. Each of those is a separate data pipeline. Separate API keys. Separate parsing logic. Separate domain-specific models. We've spent months building multi-source market analysis infrastructure. It doesn't fit in a weekend project on a $4.50 VPS. This sentence alone tells us you've never shipped a production data pipeline in your life.
"Built in Rust for speed"
-Speed is irrelevant on Polymarket. There's no order book to front-run. Markets resolve in days and weeks. This isn't Nasdaq. You didn't need Rust. You needed a better story.
"If balance hits $0, the agent dies. So it learned to survive."
-This is the part that's actually offensive to anyone who works in AI. A Kelly criterion sizer with a 6% max doesn't "learn to survive." It follows a formula. There's no reinforcement learning loop described. There's no training. There's no evolution. You bolted a risk parameter onto a script and wrote it like it's sentient. It's not.
And the $2,980 screenshot? Where is it. Show the Polymarket transaction history. Show the wallet. Show the P&L curve. Show the API logs. Show anything.
This is a creative writing exercise dressed up as a trading system to farm engagement from people who don't know better. We built Pelican for people to know better. That's why we're commenting.
We build AI trading tools. Every day. With real data, real backtests, real costs, and real users. This post disrespects everyone doing the actual work.
You didn't build a surviving AI trader. You built a Twitter thread. 🦩
Jack retweeted
Adding some new features this week 👀 Always looking to get better, love feedback. And yes, my video editing isn't sexy, but neither is launching a startup. @PelicanAI_
#fintech #trading #startup #startuplife #AI
Jack retweeted
Considering buying this dip in $AMD?
AMD has dropped 10%+ seventeen times in the last 5 years.
What happened in the 5 days after each crash?
11 of 17 bounced positive (65%)
Average return: +2.6%
Median: +1.3%
Best recovery: +21.8%
Worst follow-through: -9.3%
Not a signal. Just the receipts. @PelicanAI_
Jack retweeted
You can get this exact type of data with Pelican.
Prompt used:
Analyze AMD tick data from this morning (February 4, 2026) around market open. AMD reported earnings and is down 14%. What did the NBBO spread and order flow reveal during the first hour of trading? Was this panic selling or orderly distribution? Where did buyers step in?
$AMD Earnings Tick Data Analysis
Time: Feb 4, 2026 - First hour after market open
AMD down 14% on Q4 earnings
NBBO Spread Analysis:
• Average spread: $1.25 (59.48 bps)
• Spread volatility: 39.61 bps
• Max spread: 317.33 bps (5x the average)
• Min spread: 22.18 bps
• 89 bars analyzed
Verdict: PANIC SELLING, not orderly distribution.
Spreads blew out to 317 bps at peak, characteristic of fear-driven selling. Orderly distribution would show narrower, more stable spreads.
Buyers stepped in when spreads narrowed to 22 bps. Watch for spread stabilization as a potential buying signal.
Really proud of the work that was put into this. People will find amazing value in Pelican!
Pelican is officially open.
Traders ask questions every day like:
Why did this move?
What usually happens after this event?
Does this pattern actually work?
Pelican answers with data, not opinions.
10+ years of market history. Real context. Real evidence.
Happy trading.
1
60
Jack retweeted
Pelican is officially open.
Traders ask questions every day like:
Why did this move?
What usually happens after this event?
Does this pattern actually work?
Pelican answers with data, not opinions.
10+ years of market history. Real context. Real evidence.
Happy trading.
Jack retweeted
Big moves get attention. What comes next gets ignored.
This walkthrough covers:
• Intel’s +10% weekly gaps (14 total)
• What history actually says about follow-through
• Mag 7 risk building under the surface
• Notable market movers from today’s session
Charts on screen. No scripts. Using Pelican live.
Jack retweeted
Stats > Opinions.
Looking at Taiwan Semiconductor Manufacturing Company $TSM through an event-study lens tells a clear story.
-9 of the last 10 earnings: TSM was negative the following day
-Only 1 of the last 5 earnings led to a positive return one month later
That’s not sentiment - that’s pattern.
Since Earnings printed last week, Total return: -3.38%
Despite a couple of green days, downside pressure dominated. The worst session saw a -4.63% drop, overwhelming the short-term bounces.
Takeaway:
TSM earnings haven’t been about upside surprise - they’ve been about risk containment. The data favors caution, not conviction.
This is why data matters more than narratives.
#TSM #Earnings #Trading #EventStudy #MarketData
Jack retweeted
🟩US stocks above $100 with notable moves today🟩
Memory chips are running:
- $MUU +15.3% (2x Micron ETF) - $100B NY Megafab groundbreaking
- $RMBS +14.7% - DDR5/data center demand ahead of Feb earnings
- $SNDK +9.8% on 15.8M shares - NAND shortage driving AI storage boom
Defense moving too:
- $TDY +9.6% at $620 - Q4 earnings today, just won $32M Stryker contract
All near session highs. Info from one question.
#FinTwit #StockMarket #Trading #Semiconductors #AI #NAND #MemoryChips #TradingTools #MarketAnalysis #Stocks
Jack retweeted
$NFLX Earnings after the close today.
What 10 quarters of data shows:
• 5 up, 5 down — pure coin flip
• But when it moves, it MOVES (avg ±8%)
• Direction tends to hold into week 1
Not a prediction. Just the pattern.
#NFLX #Earnings #OptionsTrading #FinTwit #StockMarket $Netflix #Netflix #StocksToWatch #StocksInFocus
Jack retweeted
Are you new to trading and never quite sure if what you’re drawing on the chart actually means anything?
With Pelican, you can literally drop in your chart or drawings and ask:
“what is this?”
It’ll tell you what pattern you’re looking at, what it usually implies, and what to watch next...without pretending it can predict the future.
Learning charts shouldn’t feel like guessing in the dark.
#TradingBasics #NewTraders #TradingEducation #LearnToTrade #StockMarket
A lot of value to be had here!
Starting to have some fun with my new toy! (Pelican)
Equal Weighted S&P ( $RSP) down -1% today while $SPY is holding up.
Here's what Pelican caught..
PELICAN: The chart you've shared shows RSP breaking out of a rising wedge pattern. Here's what this could mean for you:
Analysis of the Rising Wedge
Rising Wedge Pattern: This pattern typically suggests a bearish reversal. The pattern forms when prices make higher highs and higher lows, but the pace of appreciation slows down, indicating weakening momentum. The break below the lower trendline confirms the bearish signal. This suggests a potential decline in price, as we see with today's drop.
Why This Matters: RSP is equal-weight S&P 500. With it breaking down while the market-cap weighted S&P 500 is holding up better suggests:
- Large-cap stocks are outperforming smaller components
- Potential rotation away from equal-weight strategies
- Broader market leadership narrowing
Trading Implications:
Target Levels: Once a wedge breaks, a common price target is the beginning of the wedge, which is roughly where the pattern started forming. Monitor potential support around the 50-day SMA around $189 and the 200-day SMA near $182.
Critical Technical Levels
IMMEDIATE SUPPORT:
$190.00 - Psychological level and first major support
$189.12 - location of 50-day moving average
$185.00-$187.00 - Prior consolidation zone
KEY RESISTANCE:
- $193.50 - Today's high and potential failed breakout level
- $195.00 - Previous wedge support turned resistance
- $197.00 - Upper wedge trendline
Risk Factors
- Broader market strength could lift RSP despite technical weakness
- Equal-weight nature makes it sensitive to small/mid-cap rotation
- Holiday season low volume could create false signals
Your rising wedge identification is spot-on technically. The pattern is playing out as expected with today's breakdown. Watch for volume confirmation and how it handles the $190 psychological level.
@PelicanAI_
#TechnicalAnalysis $SPY #MarketAnalysis #Trading #SPY #AI
21
Jack retweeted
Starting to have some fun with my new toy! (Pelican)
Equal Weighted S&P ( $RSP) down -1% today while $SPY is holding up.
Here's what Pelican caught..
PELICAN: The chart you've shared shows RSP breaking out of a rising wedge pattern. Here's what this could mean for you:
Analysis of the Rising Wedge
Rising Wedge Pattern: This pattern typically suggests a bearish reversal. The pattern forms when prices make higher highs and higher lows, but the pace of appreciation slows down, indicating weakening momentum. The break below the lower trendline confirms the bearish signal. This suggests a potential decline in price, as we see with today's drop.
Why This Matters: RSP is equal-weight S&P 500. With it breaking down while the market-cap weighted S&P 500 is holding up better suggests:
- Large-cap stocks are outperforming smaller components
- Potential rotation away from equal-weight strategies
- Broader market leadership narrowing
Trading Implications:
Target Levels: Once a wedge breaks, a common price target is the beginning of the wedge, which is roughly where the pattern started forming. Monitor potential support around the 50-day SMA around $189 and the 200-day SMA near $182.
Critical Technical Levels
IMMEDIATE SUPPORT:
$190.00 - Psychological level and first major support
$189.12 - location of 50-day moving average
$185.00-$187.00 - Prior consolidation zone
KEY RESISTANCE:
- $193.50 - Today's high and potential failed breakout level
- $195.00 - Previous wedge support turned resistance
- $197.00 - Upper wedge trendline
Risk Factors
- Broader market strength could lift RSP despite technical weakness
- Equal-weight nature makes it sensitive to small/mid-cap rotation
- Holiday season low volume could create false signals
Your rising wedge identification is spot-on technically. The pattern is playing out as expected with today's breakdown. Watch for volume confirmation and how it handles the $190 psychological level.
@PelicanAI_
#TechnicalAnalysis $SPY #MarketAnalysis #Trading #SPY #AI
Jack retweeted
NVIDIA reports earnings today. Here’s the truth about what actually happens after NVDA earnings.
Pelican analyzed NVDA’s last 9 earnings events (roughly the past 2 years). Recently, NVDA earnings have become a coin flip that gets worse the longer you hold. You’ve got a 56% chance of being green the next day… but only a 33% chance after a month.
Averages
• 1-Day: +1.28%
• 1-Week: –0.03%
• 1-Month: +3.92%
Win Rates
• 1-Day: 56%
• 1-Week: 44%
• 1-Month: 33%
The stats that matter:
• Only 3 of the last 9 earnings delivered positive 1-month returns.
• Feb/May 2024 were outlier wins (+35% and +24%) during peak AI mania.
• Since then? Only 1 of the last 5 stayed positive after 30 days.
The pattern:
NVDA earnings are violent moves that often fade.
Quick exits usually outperform diamond hands.
Range of outcomes:
• 1-Day: –8.48% → +16.40%
• 1-Month: –15.12% → +35.52%
If you want Pelican AI to run this same study on any ticker, just ask in the comments:
“Run an earnings event study on TICKER.”
What’s your $NVDA earnings play? 👇
Jack retweeted
META ANALYSIS — November 2025
CURRENT SITUATION:
$META is down -16% to $609 after Q3 earnings (Oct 29, 2025).
THE NUMBERS:
-Revenue: $51.24B vs $49.51B est. (+3.5% beat)
-EPS: $1.05 vs $6.72 est. (-84.4% miss)
-Daily active users: 3.54B (+8%)
-Ads: +14% impressions, +10% pricing
HISTORICAL PATTERN:
Pelican back tested every -15% META drop within 2 weeks since the 2022 ticker change.
5 instances:
Sep 12 → Sep 26: Bottom $136.37 (-19.29%)
Oct 20 → Nov 3: Bottom $88.91 (-32.40%)
Apr 11 → Apr 25: Bottom $441.38 (-15.63%)
Mar 25 → Apr 8: Bottom $510.45 (-18.50%)
Oct 29 → Nov 12: Bottom $609.01 (-18.98%)
Stats:
100% win rate (4/4 completed)
Average 1-month return: +16.01%
Range: +0.84% → +37.70%
THE CRITICAL DIFFERENCE:
Past META crashes: sentiment-driven, fundamentals intact
This META crash: a real earnings miss + capex reality check
Nov 2022 ($88): peak irrational fear → massive bounce
Nov 2025 ($609): rational concern about margins
THE QUESTION:
Is it time to buy $META?
#META #FACEBOOK $FB #fallingknife
Jack retweeted
We built an AI model that every trader needs. From beginner questions to deep market analysis.
Pelican speaks anyone's language.
"What is a stock?"
"Backtest a daily SPY long-only 10/30 SMA crossover from March 1st 2022 to March 1st 2023, going long when the 10-day SMA crosses above the 30-day SMA and exiting when it crosses back below"
This is Pelican. Your edge, amplified.
Launching soon.
Jack retweeted
Hey all, I have something I’m finally ready to share.
I’ve been quietly building Pelican AI, a trading assistant that meets you exactly where you are.
Whether you’re asking “What’s a moving average?” or running complex backtests like “Show every session where Microsoft’s intraday high was 2.5+ standard deviations above the open,” it’s the same interface, same speed, same clarity.
With Friday’s sharp drop (likely tariff-related), I asked Pelican to analyze how the QQQ has historically behaved after similar 2025 declines. Suggesting this decline may continue. I focused on this year for relevance, but Pelican can analyze years of market history just as easily.
What normally takes days of coding and thousands in data fees took Pelican under 30 seconds:
• 5 matching events
• –1.26% average two-week return
• Clean visualizations
• Smarter trading decisions
(Attaching the chart below)
Tomorrow, someone else might simply be learning how support and resistance work, and both questions will get the same level of care and precision.
I’m building Pelican alongside Ray Campbell, who helped architect the NYSE ARCA system — one of the engines that keeps global markets running today. His experience gives Pelican the same reliability and precision traders expect from the markets themselves.
This isn’t about building another “AI trading bot.”
It’s about giving everyone, from students to professionals, access to the same analytical firepower hedge funds use.
Your intuition, now backed by institutional-grade analytics.
Because great tools don’t raise walls; they tear them down.
I’m building the trading assistant I wish existed when I started, and one I’m still excited to use today.
The goal is simple: equip every trader (or anyone curious) with tools they’ve never had access to before.
Follow @PelicanAI_ — we’re about to start answering real trading questions across X and Reddit, dropping market stats like these, and showing what’s possible when AI actually understands traders.
Launch coming soon.
Let’s see where this goes together.
If you want early access, drop your email, phone, or X handle (no cost) at pelicantrading.ai — we’ll reach out once, no spam.














