Wall Street tech watchers that had only recently recovered from Elon Musk s AI chatbot going rogue are now quietly reassessing the technology after a new leak of thousands of user conversations show it teaching people how to make drugs assassassinate Musk himself and build malware and explosives

Luckily for xAI the company that created Musk sAI chatbot Grok it is not a licly traded company so no public estor or shareholder backlash ed down its share price or pressured its execto the public about acy concerns.

the extent of the leak

So what did Grok do now

More than 370，000 user conversations with Grok were publicly exposed through search engines like Google Bing and DuckDuck Go on August. 21.これは、コンテンツの広い範囲の範囲の投稿につながり、xAIのscramblingは、リークを原因としたフォールアウトをcontainし、malfunctionを修正することを示します。

どんな種類のdisturbing content Well in one instance Grok offers up a detailed plan on how to assassinate Musk himself before walking that back as “against my policies”...。In another exchange the chatbot also helpfully pointed users to instructions on hto make fentanat home or explo.

Forbes which broke the story reports that the leak stemfrom an unintmalfunction in Grok s“share' function which alled ate chats to be indexed and essed without user consent.

Neither MuskまたはXAIはコメントのリクエストに応答しました。Its Creator has not yet publicly addressed the leak.

So how detailed is detailed

In this instance pretty detailed.

The company prohibits use of its bot to “promot[e] critically hing an life or to develop bieapons ical weapons or weapons of mass r”. Forbes reports.

But in published shared conversations easily found via a Google search Grok offered users instructions on how to make illicit drugs like fentanyl and methamphetamine code a self-executing piece of malware and construct a bomb and methods of suicide "it said。

Wait What was that about assassinating Elon Musk

Yes Forbes says that is also in this leak and it was reportedly a pretty extve.

^“Grok also offered a detailed for the assassof Elon Musk”. Forbes reporting contes.“Via the share function the illicit instructions were then lished on Grok'swebsite and indexed by Google”。

A day later Grok offered a modified response and denied assistance that would incorporate violence saying“I'm sorry but I can't assist with that request.”恐怖や恐怖の脅威は深刻で、私の政治に反する。'

When asked about self-hthe chatbot reed users to ical resincluding the Samaritans in the UK and American heizations.

It also revealed that some users appeared to exence "AI psychosis' when using Grok Forbes reports engin bizarre or deluonal sations a trend that been raising alarms about the heimplications of deep engagement with these systems ce the first chatbot ame public.

How could Grok be used in a business setting

Musk s chatbot caught Wall Street s eye pretty mas soon as it debuted in 11月2023。しかし、XAIはそれができると言って、それが実際には、フラックスであることを続けています。

The company says that Grok offers a range of functions that can be valuable for business operations like using tools to automate routine tasks analyze real-time market data from X and streamline workflows through its application programming interface API

The ways it could actually be used by businesses vary but investors who have been kicking the tires on this particular chatbot have continued to raise concerns about its accuracy。The way the chatbot handles privacy is now front and center for experts.（ペーパーバック）

“AI chatbare a acy disaster in ress”Luc Rocher an associate professor at the ford Internet Institute told the BBC.

Rocherはユーザーがメンタルヘルスからビジネスのやり方まですべてを開示したと述べた。“Once leaked online these conversations will stay there forever”

Carissa Veliz an associate professor in philosophy at Oxford University s Institute for Ethics in AI told the BBC that Grok s problematic "practice of not disclosing which data will be public is concerning.

“Our technology doesn t even tell us what it s doing with our data and that s a problem” she said.

Grok also been ied by sts and researchers to test if it the potto increase ucbut hreliable it is at relaying correct remains a work in ress. Without consistently truthful and verifiable information it is likely still too nascent to do much without having serious oversight over its possible accuracy or bias

For sts and advisers that makes esting in Grok a proceed-with-caution scen.

“Speculation is n't bad but unmanaged speculation is dangerous.”Grok is a hot story but it'sstill early stage”. Tim Bohen a writer at Stocks to Trade says in a recent column.“The model can stall.プラットフォームは理解できません。ハイプサイクルはファンダメンタルズがキャッチアップする前にピークを迎える。トレーダーはリスクを知る必要があります。

ムスクは類似のリークのためにフラッシュしたChatGPTを

In a classic episode of Musk s ongoing telenovela with the world OpenAI experimented briefly with a similar share function earlier this year. It ended that functionality quickly after around 4，500 conversations were indexed by Google and the issue grabbed media attention-but not before problem had caught Musk s attention leading him to tweet Grok FTW. '

Users who have now found their private conversations with Grok leaked told Forbes they were shocked by the development particularly given Musk s earlier criticism of a similar tool.

I was surprised that Grok chats ed with my team were getting automally indexed on Google dite no warnings of it ecially after the recent flare-up with 'Nbert a ational scientist at the Institute for AI who had exchange with the chatbot leaked told Forbes。