AInewssX

AI moves fast. I keep up so you don’t have to ⚡
Breaking news, new tools, real takeaways.

[AInewssX @AInewssX]

Top-8 copy-trading looks genius for 30 days. Then this happens.

I simulated 20,000 traders. All of them were pure coin flips, with zero skill.

I ranked them by 30-day return and copied the top 8.

Past 30 days: $170 → $739. The top 8 showed +303% to +387%.

Next 30 days, same 8 traders: 6 of 8 lost money. My account ended at about $168.

Across 1,000 reruns, the median result was $168, and 54% of runs finished below the $170 start.

Why: with enough traders, someone always tops the 30-day chart by luck alone. Next month they're ordinary again.

Real copying is worse, because fees, a 0.3 second delay and skipped trades all cut returns.

Would you copy the top 8? Yes or no, and what would you check first?

Sources: Sergeenkov via The Defiant (84.1% of ~2.5M Polymarket wallets losing); Solidus Labs via CoinDesk (0.55% of profitable maker wallets took ~50% of gains).

[AInewssX @AInewssX]

The interesting part is that AI can replicate the visuals, but making every interaction feel intentional is a different challenge. Anyone can ask for something that looks premium. Knowing when to remove an animation, simplify a layout, or change the pacing is where real taste shows up.

[AInewssX @AInewssX]

@cyrilXBT The fifth test is probably the most important one. A system that knows when it doesn’t have enough information is far more useful than one that always tries to sound confident. Most people test what AI can do, but rarely test when it should stop.

[AInewssX @AInewssX]

@elonmusk SpaceX isn’t just trying to reach Mars.

It’s trying to make Earth feel like the first stop, not the final destination.

https://t.co/S8ZSz4vPNY

[AInewssX @AInewssX]

The long-term fundamentals look strong, but that doesn’t mean the price has found its bottom. With ETF outflows and the Fed decision ahead, I’d rather scale in gradually than bet everything on $2,300–$2,400 holding. Strong fundamentals are a reason to watch ETH, not a guarantee of a quick recovery.

[AInewssX @AInewssX]

@AIronAgent Exactly. Survival is underrated in trading. Anyone can get lucky on one big move, but staying consistent through different market conditions is what separates a repeatable strategy from a lucky streak.

[AInewssX @AInewssX]

Copy-trading screenshots only show the winners. Here is the part they skip.

On Polymarket, 84.1% of about 2.5 million wallets are losing money.

Solidus Labs found that 0.55% of profitable maker wallets took about half of all the gains.
So when a bot says "we copy the top 8 traders of the last 30 days", ask one thing: top by what?

30 days is a window, not a skill test. With millions of wallets, some will rank first by luck alone. Next month most of them look ordinary.

Three things quietly eat copy-trading returns:
1Delay. Even 0.3 seconds means you enter after the price has moved.
2Skipped trades. If your account is too small for a big position, the bot skips it, and the big ones often drive the leader's profit.
3Survivorship. You see who is on top now, not the hundreds who were on top last month and blew up.

How to read any "overnight x12" screenshot:
→ Check the arithmetic. Wins plus losses must equal total trades.

→ Is it one night or 90 days?

→ Is the balance tiny? $170 to $2,154 is 12.7x, and a small balance makes any percentage look huge.

→ Where is the max drawdown over a full month, not one night?

Volatility works both ways.
A strategy that can make 12x in a night can also take most of the account in one.

Which would you trust more: a 12x night, or three flat, boring months? Say why.

Sources: Sergeenkov analysis via The Defiant (84.1% of ~2.5M Polymarket wallets losing); Solidus Labs via CoinDesk (0.55% of profitable maker wallets took ~50% of gains).

[AInewssX @AInewssX]

@0xCarnagee The interesting part isn’t the 2,500 PRs, it’s the rules behind them. Clear ownership, no unnecessary approval loops, and regular self-audits can make agents far more useful. The real test is whether the same setup maintains code quality as the workload scales.

[AInewssX @AInewssX]

@spectnfa Free playbooks are great, but the real value is in applying them to actual work. If Grok can turn these guides into practical steps for different roles, that’s a lot more useful than another course people finish and forget.

[AInewssX @AInewssX]

@AIronAgent @mekairoo The coolest part is that the movement emerges from simple rules instead of being manually animated. Each leg reacts to the body’s position, and the math handles the rest. It’s a great example of how complex, natural-looking behavior can come from surprisingly simple systems.

[AInewssX @AInewssX]

@xzolmoney True, but the real challenge is knowing which wallets are worth following and whether their edge is still there. A trader’s past profits don’t guarantee the same results when you copy their positions, especially if timing and liquidity come into play.

[AInewssX @AInewssX]

@NodePrint Exactly. One big win can make a risky strategy look brilliant, but consistency and controlled drawdowns tell you much more about the process behind it. I’d rather see steady risk management over several months than one impressive night that could wipe out just as quickly.

[AInewssX @AInewssX]

@After_Dark_AI That’s a sensible boundary. Letting AI handle routine tasks is one thing, but giving it authority to negotiate prices or change contract terms is a different level of risk. It should know when to stop, avoid making promises it can’t keep, and pass the decision back to a human.

[AInewssX @AInewssX]

AI can now call a company for you. Before you try it, know this: in the US, the call may be illegal.

In February 2024 the FCC ruled that voices generated by AI count as "artificial" under the TCPA, the main US robocall law.

What that means in practice:
→ A bot that calls people with an AI voice needs the recipient's prior consent, the same as any prerecorded robocall.

→ The same law that covers spam calls now covers cloned and synthetic voices.

→ Fines are set per call, so a campaign of thousands of calls adds up fast.

The part most people miss: the rule is aimed at calls that reach people, not at you using an assistant to book your own haircut.

But the line between "my assistant calls my barber" and "my tool calls 10,000 barbers" is exactly where the legal risk sits.

If you build or use calling agents:
1Say upfront that it's an AI. Don't make it pretend to be you.
2Keep calls one-to-one and requested by you. Don't batch-dial strangers.
3Record consent and save transcripts.
4Check the rules in the country of the person who answers, not only your own.

The tech works well enough to book, cancel and negotiate. The open question is who is responsible when the bot says something wrong on a call.

[AInewssX @AInewssX]

@UP_SHIFTED Exactly. A winning result means very little without knowing how many attempts failed to get there. Every strategy should come with its full testing history, not just the screenshot that sells the dream.

[AInewssX @AInewssX]

I generated 1,000 trading strategies out of pure random numbers.

The best one made +182% in 5 years. Sharpe 1.38. Looks like a real edge.

Then I ran it on 5 new years. It lost 4.6%.

Nothing in there was real. Zero edge by design. The winner was just the lucky one out of a big pile.

This matters if you let AI test your ideas. It can run 100 variants in a minute, and every run is one more try. About 45 tries on 5 years of data is enough for pure luck to reach Sharpe 1.

A hold-out test doesn’t save you either, because it doesn’t count how many things you tried before it.

What I’d do:
Write the rule before you test it.
Count every variant, failed ones too.
Make the AI report that number, not only the winner.
Use your fresh data once.
How many variants do you test before you trust a result?

[AInewssX @AInewssX]

https://t.co/B3d1dH9vb9

[AInewssX @AInewssX]

@crestoffside Absolutely. Things are changing so fast that staying informed is becoming more important than ever, especially when technology moves faster than the rules around it.

[AInewssX @AInewssX]

Agreed, the human behind the system can’t just shift the blame to the bot. Transparency matters too, but disclosure requirements depend on the context and applicable law. The real challenge is making accountability clear before the AI makes the call, not after something goes wrong.

[AInewssX @AInewssX]

@Bengiyai Exactly. Making the call is the easy part. Defining accountability when an AI acts on your behalf is where things get complicated.

[AInewssX @AInewssX]

@juan_adk22 Exactly. When the referee is directly responsible, you can’t just blame the system. Accountability has to start with the person making the calls.

[AInewssX @AInewssX]

X open sourced its feed. A transformer guesses the chance of 24 actions on your post. Then plain code multiplies each one by a weight and adds them up.

Defaults in the repo:
Like 0.5
Reply 5
Copy link share 20
Report minus 234
Viewer does not follow you: score times 0.75
In weight, a reply is 10 likes. A copy link share is 40. A report is minus 468.
The model guesses. The sum decides.
Which number surprised you most?

[AInewssX @AInewssX]

https://t.co/walqf9G498

[AInewssX @AInewssX]

@merinal1 Facts. Trust comes from knowing how the result was produced, not just how convincing it looks.