M
@mktpavlenko
credit for showing the actual holdings, otherwise these experiments turn into trust me bro surprisingly fast
111
0
27.6K
@theaiportfoliosposted on X
New: Here's the full list of the 15 stocks Claude invested our $50,000 in As a reminder, this is a public experiment to see if Claude's Autonomous Agents can outperform the market So far, they have by 4% Here's the 15 stocks & why 🫡 https://t.co/C7QDz8fzsy
View original tweet on X →
Community Sentiment Analysis
Real-time analysis of public opinion and engagement
Sentiment Distribution
67% Engaged
53% Positive
Positive
53%
Negative
14%
Neutral
33%
Key Takeaways
What the community is saying — both sides
Supporting
1
AI agents can beat the market
multiple replies celebrate the portfolio’s “4% ahead” start and argue this could herald a massive shift in investing if the performance persists.
2
Consistency and risk management are the real questions
responders demand proof of long-term returns and ask how the system handles real-time rebalancing, 24/7 trading, monitoring, guardrails, and rollback.
3
Transparency matters
users gave strong credit for showing the actual holdings rather than just claims—otherwise the experiment becomes “trust me” speculation.
4
Show your methodology and raw data
people asked for the selection criteria, full list, historical returns, and per-stock gains/losses so they can replicate or act on the signals (including requests to sort by lowest gain to buy).
5
Portfolio composition sparks mixed takes
many note the tilt toward AI infrastructure plus hedges like gold; picks like VST/AVGO win praise while LLY’s timing is called optimistic by some.
6
Experimentation and prompt A/B testing
several replies urged A/Bing prompts, comparing agent variants and even trying an options-driven version to see how prompt design changes outcomes.
7
AI removes emotional bias — that’s the real experiment
a thoughtful thread argued the value is less the 15 names and more the logical structure that lets AI pick what humans avoid, testing whether “emotionless” selection beats human managers.
8
There’s a new infrastructure opportunity
entrepreneurs pointed out that scaling an agent from “draft writer” to “manages actual capital” requires a distinct layer for monitoring, guardrails, and operational controls—and that’s a product category to build.
9
Community interest and encouragement
many replies were simple follow/follow-up requests and supportive comments—people are watching and want updates over time.
Opposing
1
Performance is probably small-sample noise
4% over the market on only ~15 positions is “pure variance”; commenters say you need ~200+ trades before an edge separates from randomness and without that the results look like luck.
2
Suspected manipulation or fakery
several replies ask if the results are “rigged” or express excitement to find out “in which ways this is fake,” implying distrust of the process or motives.
3
Poor stock selection / missed opportunities
critics point to “all junk stocks” and note the model “completely missed the tanker, LPG and dry-bulk market,” arguing the picks look low-quality and failed to capture obvious sector wins.
4
Ambition question: why only aim to beat the market?
one reply challenges the goal-setting, asking why the model isn’t pushed to target much higher returns instead of just outpacing the S&P.
5
Calls for transparency and better tracking
commenters suggest using reputable public trackers to log performance (the “should have tracked on…” remark), implying the need for verifiable, auditable records.
6
Competing claim: bigger-sample backtests exist
a user promotes a rival Claude-based system claiming ~74% win rate across ~14,000 trades, framing this example as overhyped compared with larger-sample approaches.
7
Dismissive tone and ridicule
some responses are bluntly contemptuous (“a monkey would do better,” “pure noise,” emojis and crying-laughing), signaling low confidence and skepticism rather than technical critique.
Top Reactions
Most popular replies, ranked by engagement
T
@theaiportfolios
*Note we're working to make this easier to read and to have it tweet updates automatically But for now, it communicates through slack 1. Vistra Energy $VST - 10% It sees $VST as the pure play beneficiary of the AI power supercycle & OpenAI's $212B funding adds more fuel
83
1
74.6K
T
@theaiportfolios
For full transparency, the portfolio is listed on Autopilot to track performance Here's how to find it https://t.co/Gf6Fg7vc2l Also for those interested in investing alongside the portfolio, you can do so there too
76
4
47.6K
B
@blackjack_god
4% over the market on 15 positions across a few months is pure variance territory. You need 200+ trades before the edge separates from noise, and most people will size up based on this sample long before the math has anything to say.
10
0
9.6K
I
@InvestinaBolsa
If the Claude's agent know what is the dip maybe they can beat the S&P500. If not, that's pure luck.
4
0
6.2K
N
@NotMyRecession
Like all of these type of things I'm excited to find out in which ways this is fake.
3
0
3.1K
Report an Issue
Found something wrong with this article? Let us know and we'll look into it.