3G·

AI experiment - IT'S OFF!

Hello everyone,

As announced, I'm starting my experiment today: Human gives AI money, AI decides, human executes blindly.

For anyone who missed the first post: I have put €1,000 of "play money" into a separate clearing account with Trade Republic. As of today (05.01.2026), Gemini 3 Pro has taken over the wheel. My influence is limited to pressing the buttons.

The general conditions:

Starting capital: €1,000

Start: 05.01.2026

Broker: Trade Republic

AI: Gemini 3 Pro

Goal: Maximum return by the end of 2026 (high risk / high reward).

Strategy: Everything is allowed - active trades, crypto, turnaround bets.

The rule: I don't argue. If the AI says "Buy XY", then I press the button. (Unless it's not technically possible).

The AI's decision (day 1)

I gave the AI the current market data this morning and asked for an allocation. The result is... aggressive. We are going "all in" on tech, space and crypto. No ETFs, no dividend aristocrats, pure belief in growth.

Here is the execution:

DeepDive:

1. Palantir Technologies ($PLTR (-5,27%)) - The geopolitical software platform

Rationale: Palantir is the operating system for modern warfare and western corporate control.
Deep Dive: In 2026, the company benefits massively from geopolitical instability. Western governments upgrade their IT systems. In the commercial sector, the Artificial Intelligence Platform (AIP) scales rapidly as it allows companies to securely apply LLMs to their own data.
Timing: Use the setback at the start of the year (-5.6 %). Despite the high valuation (P/E ratio > 300), the growth (> 40%) justifies the price.

2. Rocket Lab USA ($RKLB (-1,04%)) - The infrastructure of space

Rationale: The duopoly in space transportation alongside SpaceX.
Deep Dive: The catalyst for 2026 is the Neutron rocketwhich is aimed at mega-constellations. In addition, Rocket Lab generates stable revenues through "Space Systems" (satellite construction), which makes it more resilient.
Timing: We are buying the growth potential of the Neutron rocket before the first launch.

3. ionQ Inc ($IONQ (+1,64%)) - Betting on the next computer revolution

Rationale: Leading "pure play" in quantum computing (trapped ions technology).
Deep Dive: Set for 2026 delivery of the Tempo system (AQ 64), which beats conventional supercomputers. IonQ has a strong balance sheet (approx. USD 3.5 bn cash) to cover R&D costs.
Risk: Extreme volatility, but analysts see doubling potential.

4 SoFi Technologies ($SOFI (+2,97%)) - Profitable disruption in banking

Rationale: The biggest beneficiary of the interest rate turnaround.
Deep Dive: SoFi is now GAAP profitable. In 2026, the bank benefits twice from falling interest rates (cheaper refinancing & rising demand for credit).
Catalyst: Q4 earnings on January 30 could exceed expectations again.

5. solana ($SOL (+1,07%)) - The institutional crypto asset

Rationale: The fastest blockchain for mass adoption.
Deep Dive: Solana runs stable and enables transactions in milliseconds (important for Visa partnerships & DePIN).
Catalyst: The "GENIUS Act" and speculation on a spot ETF are driving institutional interest. We use SOL as a "high-beta" play on the crypto market.

Transparency: The prompt

Just so you know how I talk to AI(prompt is also AI generated) - here is the exact prompt I used to generate this list:

"You are acting as a professional portfolio manager for a real money experiment starting today (05.01.2026).

Your task: Perform a deep market analysis of the current situation (as of January 05, 2026). Use deep research for economic data, sector trends, crypto sentiment and geopolitical news. Create a concrete buy list based on this.

The general conditions (hard constraints):

Starting capital: Exactly €1,000.
Diversification: At least 5 positions.
Broker: Trade Republic.
Risk management: Define stop-loss prices.

Strategy: Our goal is maximum return (high risk / high reward). Look specifically for aggressive growth opportunities, 'hot topics', turnaround candidates or crypto momentum from early 2026. Stability is secondary."

Conclusion & outlook

The €1,000 has been distributed. Now it's time to wait and see.

The next update will be in 4 weeks (or sooner if a stop loss breaks or the AI wants to panic - I ask it for updates every day).

What do you think about the picks? Are we looking at a total loss with IonQ and crypto, or is the portfolio just right for the current 2026 market?

Disclaimer: The whole thing is purely for entertainment and as an experiment. This is expressly not investment advice and not an invitation to imitate. Please do not blindly burn your money.

Nuqqx

@Nuqqx

3Settimana·

Experiment: I give an AI €1,000 and do BLINDly what it says - who wants to see what happens?

Hello everyone,

At the end of the year I will receive some additional capital which I had not expected and I will divide it up - a large part will go to my son's portfolio and I have an experiment planned with part of it which I would like to document here if there is any interest.

The idea is simple: I take €1,000 of real money (in a separate gambling account) and leave the decisions entirely to an AI, in this case the new Gemini Pro version.

Important in advanceI have already mentally written off the €1,000. It is pure 'play money' for me. A total loss is therefore fully factored in and won't hurt me financially at all. That's exactly why I want to see what's possible when you replace emotions with algorithms.

The framework conditions:

Starting capital: €1,000

Start: 01.01.2026

Broker: Finanzen.net Zero

AI: Gemini 3 Pro

Target: Maximum return by the end of 2026 (high risk / high reward).

Strategy: Everything is allowed - active trades, crypto, turnaround bets.

The rule: I don't argue. If the AI says "Buy XY", then I press the button. (Unless it's not technically possible).

I have my normal portfolio for retirement provision, but here I want to know: Can the AI beat the market through active management or am I burning the €1,000?

Would you be interested in following this? I would post regular updates on the trades and performance (including screenshots). I was thinking of a 4 week cycle.

What do you think: Will this be a total loss or the way to the moon?

Disclaimer: The whole thing is purely for entertainment and as an experiment. This is expressly not investment advice and not an invitation to imitate. Please do not blindly burn your money.

222

46 Commenti

Shem_sen@Shem_sen

I'm currently celebrating and am super excited 🤩

•

PositivePossum@PositivePossum

Very nice! I'm happy to leave a follow in return.

What do you think about me defying you with GPT and trying to beat your return with the same prompt?

Because I wanted to try a similar experiment myself - but OpenAI controlled😉

•

Nuqqx@Nuqqx

@PositivePossum I'm open to it, you can always post your update under the updates :)

••

Marcel Otto@Marcel3008

Cool thing, I get a bit of FOMO 😅

•

Suky@Suky

Shall we turn it into a little game? 👀🔥
Whoever hits the yield best by the end of 2026 wins fame & glory 😄

I submit: +120 % 🚀

Let's see who's closer!

•

Jakob C.@Creutzfeldt_Jakob

@Suky Such values over a year will be impossible, I think. I would expect a maximum return of 35% after one year - which would be very ok 😅

•

Nuqqx@Nuqqx

@Suky if he keeps giving me values like that, then I'm extremely optimistic...

I was a bit afraid that it would give me really high risk junk values all the time :D

•

Suky@Suky

@Creutzfeldt_Jakob Can be good. I'm still sticking with +120%, otherwise it wouldn't be a game.
Either strong returns or we have witnessed live how an AI bubble is created! ^_^

•

Suky@Suky

@Nuqqx It is not selection alone that is decisive, but the sustainable management of risk and timing across different market phases. let's see, I'm very excited!

••

Jakob C.@Creutzfeldt_Jakob

@Suky fair enough, feel free to be so optimistic 😉 I'm keeping my fingers crossed that the result goes more in the direction of your tip than in the direction of my 35% 😅

••

Steppenwolf66@Steppenwolf66

Nice, I also used AI to get recommendations for shares, ETFs and crypto due to a lack of experience. I have already realized some gains and losses, but am mostly satisfied even if not everything went as desired. Some recommendations went straight down the drain. That's why I find your approach interesting and I'm curious to see what comes out in the end. I wish you every success.

•

Nuqqx@Nuqqx

@Steppenwolf66 I invite everyone to follow it, then you can draw experience from it and consider what you do with it!
AI is a useful tool, but I would NEVER invest blindly without looking at it myself, e.g. in a bond or long-term portfolio!

•

Steppenwolf66@Steppenwolf66

@Nuqqx I see it the same way, "can" be a useful tool. Chat Gpt unfortunately has no access to online or real-time prices, is it different for you or how do you compensate for this? For fun, I checked my portfolio today with everything except Etf and crypto with Gpt and I will possibly shift a little more, even if I am not yet so sure that I should follow the advice, as I only bought 2 positions the day before yesterday. 🤷

••

Nuqqx@Nuqqx

@Steppenwolf66 Gemini has access to the share prices, I have stored my share numbers there and it can almost always tell me the current status or the last 30 minutes.
I also have the Pro version, but I only use it for the main portfolio as a diary

•

Patrick@Piwy

Interesting idea! I am very interested in future updates.

•

bull_investor_1319@bull_investor_1319

Fingers crossed 🍀🔥

•

Sven@Nevs1848

I find it interesting.

••

Krush82@Krush82

Is there a reason why AI has chosen exactly this one?

••

Nuqqx@Nuqqx

@Krush82 yes there is, but it's incredibly long...if i post it here then every comment is gone :D

Here is the short summary of the AI:

This portfolio is an aggressive bet on US tech dominance in 2026, ignoring defensive wisdom and going all-in on the sectors that define 21st century physical and digital infrastructure: AI software (Palantir), space logistics (Rocket Lab), quantum computing (IonQ), modern banking (SoFi) and blockchain infrastructure (Solana).

We don't buy the average ("beta"), we buy the exception ("alpha"). In a world of instability, standing still is the biggest risk. With strict risk management (stop-loss) and a clear thesis, we go long into the future.

••

Krush82@Krush82

@Nuqqx It's actually more blah blah than a reason why exactly you chose them. Why, for example, are the momentum stocks micron tech, western digits, seagate etc. not included? Why don't you ask the AI? I would be really interested

••

stock_virtuoso_1968@stock_virtuoso_1968

I'm very excited about that

••

Fl_ow@Fl_ow

I'm also curious to see how it continues ^^

••

Wiktor_06@Wiktor_06

Thanks for the cool experiment. If you can use AI to build your strategy in the future and achieve positive returns, then we have found a MoneyGlitch 😂

••

D.S.@Crash-Propheteus

Great idea, then the AI takes over from the crystal ball 🔮 I'm very excited to see the results! It's like a new generation 😂

••

Christiane@CXM-AKT

Interesting selection - I'm looking forward to seeing how your experiment develops

••

Susanne@Isus01010

I am curious about the stop loss prices. In my experience, the AI cannot calculate reliably, or are they only defined as a percentage?

••

Nuqqx@Nuqqx

@Isus01010 are only defined in %, but I don't set the S/L rates either, only mental ones. It is supposed to be a high risk/high reward experiment. I carry a total loss in risk

•

Christian@Gebsen79

I also just asked Gemini how it would invest 1000 euros; I got: 400 Bitcoin, 400 Nvidia and 200 ionq....

••

diamantis@damis

I used your prompt in Gemini as well. The results are as follows.
Nvidia 25%
Solana 25%
Palantir 20%
Riot platforms 15%
Spyre Therapeutics 15%
I think every time you ask you get a different answer.

••

Nuqqx@Nuqqx

@damis Of course, there will be significantly more than five companies that meet the requirements. But the sectors are definitely similar.

•

Malthead@Malthead90

Mega cool! I can't wait to see what comes out of it! 😄

••

Fetzen@Fetzen

You can see whose AI performs better https://getqu.in/CJWwds/ https://getqu.in/CJWwds/

PS: there is no good reason to test this with real money

••

Nuqqx@Nuqqx

@Fetzen A matter of opinion. Without 'skin in the game' you don't learn anything about discipline. In the demo, everyone is Warren Buffett because losses don't hurt. But I want to know whether it works under real pressure. Apart from that, the seriousness is simply very different from paper money, but everyone is entitled to their own opinion on that.

••

Lasse@Lasse_1996

For me it was Ternium SA ADR on 01.07.2025, which open ai recommended to me in the sense of a buffet share (the 1970-1990 buffet). I bought at €26. Currently €34.60 with a 10% dividend at the time based on the share price.

••

Marc@Marcinho

Great idea. I'm excited

••

Steffen@FreigeistProtokoll

Hi... I'm new here and just came across your experiment... 3 of the assets I hold myself or are on my watchlist, so it's interesting for me...

1. Palantir (PLTR) - The only green position
Buy (05.01.): 167,65 $.
Current (06.01.): approx. 174,04 $.
Performance: approx. +3.8 %.
Status: Well above the stop loss (-15 % at approx. 142.50 $).

2. Rocket Lab (RKLB) - Strong start, then cooling off
Buy (05.01.): 71.49 $.
Current (06.01.): approx. 75.99 $.
Performance: approx. +6.3 %.
Status: Stable, stop loss is low at -20 % (approx. $ 57.20).

3. IonQ (IONQ) - Under pressure
Buy (05.01.): $46.77.
Current (06.01.): approx. 47.78 $ (temporarily below 47.20 $).
Performance: approx. +2.1 %.
Status: Just above the entry level, but volatile. The stop loss is quite generous at -25% (approx. $35).

4 SoFi (SOFI) - The shaky candidate
Buy (05.01.): 27.46 $.
Current (06.01.): Crashed to approx. 22.61 € (corresponds to approx. 24.50 - 25.00 $).
Performance: approx. -9 % to -10 %.
Status: Danger zone! The stop loss of -15 % (approx. $ 23.30) is getting dangerously close. The capital increase of 57.8 million shares is putting massive pressure on the share price, I used this as an entry point myself...

5. solana (SOL) - crypto correction
Buy (05.01.): 165,00 $.
Current (06.01.): approx. 134.34 $.
Performance: approx. -18.6 %.
Status: Critical. The stop loss is at -30 % (approx. 115 $).

At first glance, you have quite a diverse and agile group...

••

Markus@MarkusF

I ask myself, why not just make a getquin or par*** profile and enter the trades there INSTEAD of actually buying the shares?

It has 100% the same informative value.

Would also make it possible to do this with different models in parallel without getting poor.
And you could also allow her to go short, with and without leverage (which would be fundamentally stupid in the vast majority of cases with real money)

••

Nuqqx@Nuqqx

@MarkusF "A matter of opinion. Without 'Skin in the Game' you don't learn anything about discipline. In the demo, everyone is Warren Buffett because losses don't hurt. But I want to know whether it works under real pressure. Apart from that, the seriousness is simply very different from paper money, but everyone can have their own opinion on that."

••

bull_investor_2600@bull_investor_2600

10H

I can't wait to see how this develops. I'm very keen to follow it. I don't think it will be Rocket or a total loss

••

Markus G.@Novak10

A nice experiment. Feel free to update me.

••

Romainhj@romainhj

What are the sell price targets for each stock?

••

Lookas@Lookas

I'm curious and expect updates in between😁

••

Nuqqx@Nuqqx

@Lookas I'd love to, the first update will be available by the end of January at the latest

••

Chv@Chv

Cool thing, I've had gemini pro for 2 weeks, I've also been messing around with her, exported my depot and asked for her opinion, the approaches were good aaaaber when I ask for alternatives so ala why not the etf but the or the bribg I gemini quickly just to do what I question, so the ki does not directly on their indication. And recommendation... I'm curious to see how it works for you, what do you ask the AI "daily" if it wants to reallocate? or what exactly ...

••

Nuqqx@Nuqqx

@Chv Runs via a prompt that is automatically executed 3 times a day and was written by the AI itself:

Run the portfolio check based on current news and prices.

Portfolio holdings & stop-loss (SL):

Palantir (1,333 pcs) | SL: 127,53 €

Rocket Lab (3.125 shares) | SL: € 60.00

IonQ (4,651 pcs) | SL: € 32.25

SoFi (7,944 pcs) | SL: 21,40 €

Solana (1,718 pcs) | SL: 81.49 €

Task 1 (Defense):

Search for breaking news on these stocks.

Check whether stop-loss marks have been breached.

Task 2 (Offense - Hunter Mode):

Scan the market for extreme high-risk/reward opportunities that could significantly outperform our weakest position.

IMPORTANT: Only propose a SWAP if the potential profit significantly exceeds the transaction costs (order fees & spread for selling + buying new). A swap must be worthwhile 'net'.

Output: Clear instruction: HOLD, SELL or SWAP (with justification incl. cost-benefit).

•

Chv@Chv

@Nuqqx ok exciting, i'm thinking about doing the same for myself, an extra depot but i would probably put 10k right away
..the thread then works independently and reports back a task if it wants changes ? (I have gemini pro subscription) but I haven't looked into it that deeply I'm just asking questions ;-)

••