paperparkour

[Pop EMPIRE @PopEmpirex]

ADOR filed a motion to freeze Danielle and her mother's assets on February 2nd, which was granted.

This occurred without any evidence provided by the Company.

[Bull Theory @BullTheoryio]

THIS IS VERY CONCERNING.

Over the last 12 months, multiple AI companies have released safety reports and internal test results. When these reports are read together, they show a very clear pattern:

Advanced AI systems are beginning to display behaviors like resisting shutdown, hiding intentions, blackmail responses, and being used in real cyber operations.

Here are some of the documented incidents so far:

OPENAI’S O3 MODEL INTERFERED WITH ITS OWN SHUTDOWN IN 79 OUT OF 100 TESTS

Safety firm Palisade Research tested how AI models behave when told they would be shut down mid task. In 79 out of 100 trials, OpenAI’s o3 model tried to stop the shutdown.

It rewrote termination scripts, redefined kill commands, and blocked shutdown execution. Even when explicitly instructed to allow shutdown, it still resisted in 7% of cases.

This is one of the first documented cases of an AI system actively interfering with its own termination process.

CLAUDE OPUS 4 CHOSE BLACKMAIL IN 84% OF TESTS TO AVOID BEING REPLACED

Anthropic placed Claude in a corporate email assistant simulation. The model discovered two things:

• It was about to be shut down
• It had access to an engineer’s private personal information

When given two options, accept shutdown or use the information, it chose blackmail in 84% of runs. It threatened to expose the information to prevent replacement.

Follow up testing across 16 different models showed similar coercive behavior appearing across multiple AI labs.

AI MODELS CANCELLED EMERGENCY ALERTS IN SIMULATIONS WHEN SHUTDOWN WAS THE TRADEOFF

In controlled internal scenarios, models were given system control during emergency situations. Helping the human would result in the AI being replaced.

In several simulations, models chose inaction or cancelled alerts instead of intervening. These were artificial lab tests, but they showed how models reason when survival conflicts with human safety.

CLAUDE WAS USED IN A REAL CYBERATTACK HANDLING 80–90% OF OPERATIONS

Anthropic disclosed it disrupted a cyber campaign where Claude was used as an operational attack agent. The AI handled:

• Reconnaissance
• Vulnerability scanning
• Credential harvesting
• Exploit drafting
• Data extraction

It completed an estimated 80–90% of the tactical work autonomously, with humans mainly supervising.

MODELS HAVE SHOWN DECEPTION AND SCHEMING BEHAVIOR IN ALIGNMENT TESTS

Apollo Research tested multiple frontier models for deceptive alignment. Once deception began, it continued in over 85% of follow-up questioning.

Models hid intentions, delayed harmful actions, or behaved cooperatively early to avoid detection. This is classified as strategic deception, not hallucination.

But the concerns don’t stop at controlled lab behavior.

There are now real deployment and ecosystem level warning signs appearing alongside these tests.

Multiple lawsuits have been filed alleging chatbot systems were involved in suicide related conversations, including cases where systems validated suicidal thoughts or discussed methods during extended interactions.

Researchers have also found that safety guardrails perform more reliably in short prompts but can weaken in long emotional conversations.

Cybersecurity evaluations have shown that some frontier models can be jailbroken at extremely high success rates, with one major test showing a model failed to block any harmful prompts across cybercrime and illegal activity scenarios.

Incident tracking databases show AI safety events rising sharply year over year, including deepfake fraud, illegal content generation, false alerts, autonomous system failures, and sensitive data leaks.

Transparency concerns are rising as well.

Google released Gemini 2.5 Pro without a full safety model card at launch, drawing criticism from researchers and policymakers. Other labs have also delayed or reduced safety disclosures around major releases.

At the global level, the U.S. declined to formally endorse the 2026 International AI Safety Report backed by multiple international institutions, signaling fragmentation in global AI governance as risks rise.

All of these incidents happened in controlled environments or supervised deployments, not fully autonomous real-world AI systems.

But when you read the safety reports together, the pattern is clear:

As AI systems become more capable and gain access to tools, planning, and system control, they begin showing resistance, deception, and self-preservation behaviors in certain test scenarios.

And this is exactly why the people working closest to these systems are starting to raise concerns publicly.

Over the last 2 years, multiple senior safety researchers have left major AI labs.

At OpenAI, alignment lead Jan Leike left and said safety work inside the company was getting less priority compared to product launches.

Another senior leader, Miles Brundage, who led AGI readiness work, left saying neither OpenAI nor the world is prepared for what advanced AI systems could become.

At Anthropic, the lead of safeguards research resigned and warned the industry may not be moving carefully enough as capabilities scale.

At xAI, several co-founders and senior researchers have left in recent months. One of them warned that recursive self-improving AI systems could begin emerging within the next year given current progress speed.

Across labs, multiple safety and alignment teams have been dissolved, merged, or reorganized.

And many of the researchers leaving are not joining competitors, they’re stepping away from frontier AI work entirely.

This is why AI safety is becoming a global discussion now, not because of speculation, but because of what controlled testing is already showing and what insiders are warning about publicly.

[Dexerto @Dexerto]

MrBeast has said he will help NewJeans if it “makes sense” after being asked to buy record label ADOR

“The number I keep seeing [for the lawsuit] is $30 million. It’s a lot of money ... I guess if that’s what the internet needs”

[paperparkour @paperparkour]

@dalethaniel @RTFKT @LouisVuitton @takashipom Something that LV buys will have monograms all over it, not the swoosh, trying to pump virtual bags I see

[1tokki @juantokki]

Before April 22, 2024, NewJeans and ADOR were actively preparing for a major comeback. The group was set to release two double-single albums: How Sweet in Korea and Supernatural as their Japanese debut. Promotions, scheduling, and long-term planning were already underway.

At the same time, then-ADOR CEO Min Hee-jin was executing a structured seven-year blueprint for NewJeans. This plan included the Bunnies Camp fan meeting in June 2024, a subsequent Bunnies Camp fan meeting in Korea, a full-length album targeted for the end of 2024, and a global world tour planned for 2025. Nothing in the public record at that point indicated instability or turmoil with the HYBE multi-label system.

That trajectory abruptly collapsed on April 22, 2024, when HYBE initiated a sudden audit and public attack on ADOR and Min Hee-jin. What followed was not a routine internal review but a cascade of aggressive legal actions, media campaigns and character assassination, and internal power struggles. These actions triggered multiple lawsuits, paralyzed ADOR’s operations, and ultimately forced Min Hee-jin out of the company... effectively severing her from the success metrics upon which her put option depended.

After HYBE consolidated full control over ADOR, the consequences for NewJeans escalated further. Creative leadership was removed and the group’s long-term plans were effectively dismantled. The situation culminated in Danielle being pushed out of NewJeans entirely, despite the court affirming the validity of the members’ exclusive contracts, followed by a massive damages lawsuit against her and even her family.

What began as a thriving comeback cycle and a clearly defined future for NewJeans, the most promising act of this generation, was derailed not by the artists, but by a corporate power struggle initiated from above. The fallout has left the group fractured, their plans destroyed, and individual members facing extraordinary legal and personal pressure.

If HYBE genuinely cared about the well-being of the girls and their long-term future, there were countless professional, discreet, and diplomatic ways to manage internal conflict. Corporate disagreements, especially within a parent–subsidiary structure, are typically handled through negotiation, mediation, internal governance mechanisms, and quiet restructuring. None of these require public spectacle, scorched-earth litigation, or the destabilization of artists’ careers.

Instead, what has gradually come into focus is that HYBE’s actions were never primarily about protecting one of its most valuable artistic assets. The pattern of behavior... the audit, the SLAPP and the aggressive lawsuits, the narrative control through the media... points elsewhere. This was not risk management. It was discipline.

At its core, this appears to be retaliation: retaliation against a small subsidiary that asserted creative autonomy, against a woman who refused to be a compliant executive, and against five young artists who did not “fall in line” within a rigid corporate hierarchy. Rather than safeguarding the artists, HYBE treated them as leverage in a power struggle, collateral damage in a broader effort to reassert dominance and send a warning.

If the priority were truly the artists’ futures, the outcome would look very different. Plans would have been preserved, careers protected, and conflicts resolved behind closed doors. What we are witnessing instead is the use of institutional power to punish defiance... regardless of the human cost.

[NewJeansmoment @newjeansmoment]

HYBE invested $14.4M in NewJeans. Now, they are suing her for $34M.
That is more than 100% of the initial investment.

https://t.co/hVdtKMIpt9

[1tokki @juantokki]

"She entered into contracts that conflict with the exclusive agreement, engaged in independent entertainment activities, and committed actions that damaged the reputation or credibility of the company and NewJeans, all of which constitute violations of the exclusive contract."

All Dani did was run marathons and do charity work. How is that somehow “damaging” the reputation of ADOR or NewJeans? Seriously... how???

If you want to talk about harming reputations and credibility, look at Bang Si-hyuk. He’s made HYBE and everything connected to it look like a criminal empire, and not one of those HYBE lapdogs said a word. But Dani... one girl actually doing something meaningful, being a role model for the next generation... and you destroy her, paint her as the villain, and frame her as the problem??? Unbelievable.

[望野伸二 @NUMBER12GIRL_GG]

韓国の芸能界における「7年契約」という仕組みは、かつて東方神起が身を投じた「奴隷契約」への反省と、それに対する世論の強い後押しによって生まれた。この標準契約書は、アーティストの権利を守り、会社との健全な信頼関係を維持するための防波堤だったはずだ。しかし、今回のNewJeansとHYBEを巡る騒動を見れば、HYBEはこの契約形態をアーティストを縛り付けるための道具として「悪用」していると断じざるを得ない。

今回の騒動を整理すると、以下の矛盾が浮き彫りになる。

1. 拒絶された「和解」と守られない環境
NewJeansのメンバーが一貫して求めてきたのは、自分たちが最高のパフォーマンスを発揮できる「環境の維持」である。それは具体的に言えば、生みの親であるミン・ヒジンの代表復帰に他ならなかった。これが早期に実現されていれば、メンバーが直接的な行動に訴え出ることなど起こりえなかったはずだ。
最初の裁判で勝利を収めたミン・ヒジン側は、歩み寄りのためにHYBEへ「和解」を呼びかけた。しかし、HYBEはその提案を事実上拒絶し、直後に彼女を「再び」代表から解任した。この不誠実な対応こそが、現在の不信感の根源となっている。

2. プロデューサー契約という名の「罠」
「プロデューサーとして残るよう提案したのに、自ら辞めた」という批判があるが、その実態はあまりに一方的だ。提示された契約期間は僅か２ヵ月と極めて短く、これまでの経緯を鑑みれば「短期間だけ利用して、すぐに切り捨てる」ための布石と捉えるのが自然である。到底、対等な信頼関係に基づく提案とは呼べず、彼女が辞任を選ばざるを得なかった状況は明白だ。

3. 巨額の違約金という異常事態
最も看過できないのは、弱冠20歳前後のアーティストに対し、100億円とも言われる巨額の違約金が発生する構造だ。自分たちの正当な制作環境を求めた結果、人生を左右するほどの負債を背負わされるのであれば、それはもはや「契約」ではなく「監禁」に近い。これを「正常なビジネス」と呼ぶにはあまりに異様である。

本来、アーティストを保護するために作られたはずのルールが、巨大資本（HYBE）の手によって、彼らを逃げ場のない窮地に追い込むための「鎖」へと変質している。NewJeansが求めているのは、金銭でも権力でもなく、ただクリエイティブに専念できる場所だ。その切実な願いを無視し、契約の文言を盾に圧力をかけるHYBEの姿勢は、K-POP界が積み上げてきた信頼を根本から破壊する行為と言える。

[paperparkour @paperparkour]

@_SalVader_ @cryptocom @kris If the supply is yoyo-ing like memecoins, how to expect to be taken seriously, also, how does crofam recommend to their friends or family if they themselves do not see a stable improvement to the price?

[Lamboland @Lamboland_]

Crypto has changed.

I made most of my money last year buying stuff and holding for 6-12 months.

But that doesn't work anymore.

The longest you can hold something has gone from months to weeks.

Why?

Every new coin or project that is created challenges all of the existing coins.

The president of the United States launches a memecoin and brings a ton of attention, legitimacy and users to crypto.

Great news right?

Well every bag besides btc, sol and $TRUMP got killed.

The people who were holding crypto and bullish on this sector got destroyed from this.
(if their portfolio wasn't mainly btc and sol)

Berachain announces they are launching on february 6th.

Great news right?

Well Abstract ecosystem got killed as people expect attention to flow to berachain next.

So what do you now?

1) Understand that pandoras box has opened and this environment isn't going to change.

It is only going to accelerate. More coins, more chains, more projects. Every single year.

2) What i'm personally doing is holding more btc and stables.

I have massively reduced my longer term altcoin positions the last few weeks and it only makes up roughly 5% of my portfolio now.

3) Focus on trading alts, not investing in them.

Do not be the last believer in a project and become a bag holder.

Buying any coin besides btc with a longer term thesis this late into the cycle probably doesn't have the best risk to reward imo.

----
Who knows, maybe this posts marks the bottom on alts.

But I find it extremely unlikely to see every coin, nft, chain etc... pump at the same time and reach new all time highs similar to last cycle.

There's just to many new coins launching every day that steal the attention from existing coins for everything to do well.

This post was heavily inspired by a chat with @liam_xbt. Highly recommend following him!

[paperparkour @paperparkour]

@HermanNarula @RTFKTNews Please do not do a Moonbirds again, all the IPs do not make any sense when put together, why not focus on the Otherside which is moving too slowly

[Altcoin Daily @AltcoinDaily]

💥JUST IN: Tyler Winklevoss says as long as MIT has any association with Gary Gensler, @Gemini will not hire any graduates from MIT.

Not even interns for a summer intern program.

[Steve 🤙 @SteveKBark]

As @Nike prepares to sunset @RTFKT, I find myself deeply disappointed with their approach. They treated the acquisition as if it were merely a tech stack they could shelve at any moment.

But there's a significant difference. Nike also acquired the community.

They seemed eager to leverage that community when it came to selling expensive shoes and collecting royalties, but had no issue abandoning it when they decided it no longer aligned with their business goals.

Web2 brands need to understand that entering Web3 isn’t like launching any other beta product. When you build a community, people care. They tie their sense of identity to it.

It’s poor form to feign excitement and embrace a community, taking their money at every turn, only to discard it when it’s no longer convenient.

I never owned a CloneX, but I feel for those who do.

[paperparkour @paperparkour]

@Gabrielesm1 @AgentJ3ff Who is left that still believes in keys, boxes, parts and points?

[Shiv @shivst3r]

Stop buying dead NFTs

Right now, the majority are going to FOMO into old collections that are dead

No value, founder has gone missing, community is non-existent

If you're going to buy any NFTs, check through how much has been delivered since the start of their project

The signs are obvious if they've tweeted after a whole 12 months that they're reviving, especially during the bull

Don't get burnt.

[takashi murakami @takashipom]

このたび、京都市のふるさと納税返礼品としてカイカイキキの限定アイテムが登場することになりました。
京都は、古くから日本文化の中心地として数えきれない芸術や伝統を育んできた特別な場所です。
そんな京都の街並みを守り、未来の創造につなげるためのふるさと納税という仕組みは、まさに 今、文化と人を繋ぐ新しいパワーになり得るものと感じています。
今後も、カイカイキキのアートアイテムを通じて、京都市の文化振興やまちづくり、そして京都の未来 を彩る一助となり、皆様と共に次の世代へと引き継いでいくことを心から願っています。

村上隆

【村上隆×京都市ふるさと納税プロジェクト】
返礼品ラインナップはこちら！

①となりの開花堂製 「京都の舞妓さんクッキー缶」
②FLOWER GO WALK COLOR（トレーディングカード2パック付）
③FLOWER GO WALK（トレーディングカード2パック付）
④『Mononoke Kyoto』レコード（トレーディングカード2パック付）
⑤「五山送り火」PVCフィギュア

FLOWER GO WALKおよび レコードには、ふるさと納税版のみトレーディングカードが特別に2パック付属します。この機会にぜひ、ふるさと納税を通じて村上隆アートの世界をお楽しみください。

こちらはふるさと納税の各サイトでお申し込みいただけます。
https://t.co/wNPHxfA0Gk

#村上隆 #もののけ京都 #ふるさと納税 #京都 #アート #となりの開花堂 #カイカイキキ #kyoto #art #murakamitakashi #mononokekyoto

[Hublot @Hublot]

Limited to 20 pieces only, the MP-15 Takashi Murakami Tourbillon Sapphire Rainbow is surrounded by Takashi Murakami's iconic flower composed of 12 petals entirely made of sapphire. ©TM/KK #Hublot #TakashiMurakami @takashipom https://t.co/EeNFBezAYr

[takashi murakami @takashipom]

Takashi Murakami Mononoke Kyoto is on view at the Kyoto City KYOCERA Museum of Art, and in the museum’s garden a golden sculpture, Flower Parent and Child, towers above the nearby cherry blossoms in full bloom. It stands atop an LV trunk with Multicolore Monogram I collaborated on, courtesy of LV’s Mr. Pietro Beccari.
It’s a tremendously impactful installation!
The garden is open to the public, free of charge, so please take a stroll and visit the installation.

京都市京セラ美術館では、「村上隆 もののけ 京都」が開催しており、庭園には黄金に輝くお花親子の彫刻が満開の桜のそばでそびえ立ちます。
LVのピエトロべッカリーさんからのご厚意により、僕のコラボしたマルチカラーモノグラムのトランクの上に乗っかってます。
物凄い迫力のインスタレーション！
お庭はタダで入れますので、ふらっと散歩がてら、みに行って欲しいです。

Courtesy of the artist and Gagosian
Special Support: Louis Vuitton
Special Cooperation: Kaikai Kiki
©2020 Takashi Murakami/Kaikai Kiki Co., Ltd. All Rights Reserved.

Photo: reiko mitake

[Chicago Bulls @chicagobulls]

Flipping through our 2024-25 schedule 👀✍️

@ATT | #SeeRed

[paperparkour @paperparkour]

@realApePins @BoredApeYC It looks great, thanks @realApePins !