Cognit AI的动态

You always do your mistakes and then terminate people's accounts! First they were arrogant saying there were some abusive words in the job posting, while we escalate, now this is the message! Stop using such AI LinkedIn LinkedIn Help

Saw a call come in from HI. I didn't answer. No msg. Same # called again approx 10 seconds later. Didn't answer; no msg. Googled # and found web sites saying possible scam, in which scammers will say "can you hear me?" so they can record you saying 'yes'. I have had a few clients previously tell me of this scam and bushed it off contributing to media paranoia, but now it hits home. With AI, are we going to be afraid of even talking on the phone? Will we be afraid of AI duplicating peoples voices not for just one word, but unlimited words and phrases, then mimic this to targeted people's phone #s? Is this what the future holds? Will we lose our human interaction altogether? God help us all.

This is a fascinating and instructive case study of the dangers of AI, algorithms and our increasingly digital world. AI programmed bots may never be able to ape human morality, yet we are increasingly relying on them to play the role of humans. Thus, people are seeking to meet their social needs through machines that are not able to make decisions about what is good for humans in a particular predicament. If machines merely reflect what we want to hear, they are likely to act as amplifiers of existing fears, doubts and prejudices and with Big Tech platforms being solely focused on monetisation that is only likely to get worse. https://lnkd.in/ejyW6CXU

700000 views nerly 8 000 comments and 15000 likes on TikTok. My first attempt at AI. 2 years ago . Interestingly the comments go from ???? to “ the police will get you” you have no right being in SG, boring, it’s fake to “when this happen” “ was mrt late” “how to claim insurance “. With the monsoon season in SG there has been a resurgence of views and comments. It’s a very interesting excercise which proves the simple things can have good organic reach. We are working on another very cool AI project will share soon. What’s your thoughts on this example? Jason Wong Flame Rozario #AI #MRTsingapore

Well this was an extremely COOL way to use #GenAI !! I got a suspicious email to my mum's business account, and I knew it was a scam, but wondered if AI could indeed help identify it .. AND IT DID ?? ?? ?? Thanks Anthropic and #Claude What does this mean? ?? We can EASILY train vulnerable groups (older and less tech savvy) to, before clicking on any link, run it through a #GenAI to see if it's real or fake - we can run emails, email addresses, links, videos etc ... Every day over 1million get scammed ?? ??♀? This could, in a small way, help reduce those numbers Want more AI tips? Stay tuned #simpleAI #UsefulAi

Watch out for Artificial Intelligence ?????? scams Scammers use AI to clone voices ??? impersonating people. One of the scam’s using AI is where they clone voices of loved ones, then call ?? using that voice, pretending to be desperate for money ?? to get out of trouble or danger. It's very convincing ?? Learn more now ?? https://bit.ly/4e0QqHr #scams #scam #scammers #scammer #scamalert #socialmedia #aiscams

You know what, this one is personal. But it belongs on LinkedIn, Facebook and anywhere else that discussions about AI are taking root. Not because the advertising and marketing industries are dying, my mate is. And there’s a chance he shouldn’t be. Before Wunderman Thompson was absorbed into VML I had the opportunity to work on case studies for DM that highlighted the advancement of AI in medicine. Some specialist hospitals have access to AI systems that compare 1000s of MRIs and respond in minutes, based on recorded results of both survival and cause of death. It can flag things like ‘97% of the time that shadow is this, but the other 3% it’s that, which is misdiagnosed - might wanna check that out’. My friend didn’t get that. And rather than pour investment into making that kind of AI accessible, we’ve got some not-quite-Pixar middle-aged hipster on a mountain with no service in a Vodafone ad (probably the same bloke driving the Coca-Cola lorry). There’s clear IP theft and an inability to spell for some Bombay Sapphire spec work, and homogenous, plagiarised verbiage for anyone that wants to have ‘content’ rather than copy. Part of the madness is that using it where it’s needed most, won’t put doctors out of work! It’ll give them a chance to catch up. While relying on it too much in creative industries is making things worse. LinkedIn is asking me if I want to rewrite this with AI, so, if I don’t reply to your responses it’s because I’ve thrown my phone out the window again.

Unfortunately, stories like this have become far too common. Creating an AI that is able to build trust and connection with users can certainly be valuable, but not when clear safeguards are not put in place. It is beyond imperative that AI companies address these clear gaps in their models, and there is better education on the limitations of AI. Personally, this is something we take incredibly seriously at Thrive, and has informed our own strategy for AI development and integration, putting human safety at the absolute forefront. If you'd like to learn more about our strategy, DM me and I'll share our recent ThriveAI whitepaper. The key: Solving these issues in AI takes a commitment to place people above all else. #AI #mentalhealth #AIsafety

AI + Scamming = A Problem Imagine AI being used to scam you. In the context of the Clip Andrew Rohm is talking about how someone could potentially scrape, grab all of the information on, your public profile and create a almost real version of you to scam. Worst part, grandma is going to have no clue that it's not you. Lynette Bruecker Arnhart, PhD, CBCP from team Logic IT in Mosinee agreed with the statement and added on that she's trying to find ways to use it in a safe way. The problem with AI is that it's too good at its job, convincing your that it's real. Some of the scamming uses I've thought of for AI: → Relationship Modeling (Family Member or Otherwise) → Mass Scam Outreach → Personalized Scam Outreach What uses of AI should we be scared of? --- P.S. I'm a Digital Marketer If you want to learn more about Digital Marketing Subscribe to my Newsletter. Links in my Bio :) or go to: https://lnkd.in/gchYvCp7

???????????????????? ???????????????????? ???????????????????????? - ?????? ???????????? ???? ???????????????????????? ???????????????? ???????????? AI is a powerful tool that can be utilized in various ways, including potentially aiding criminals in their activities. While platforms like ChatGPT and DALL-E have safeguards in place to prevent misuse, there is still a risk of manipulation for illicit purposes. However, it is crucial to acknowledge that AI has numerous beneficial applications beyond criminal intent. As society becomes more familiar with AI technology, the initial fear and resistance will likely diminish, much like past concerns about new media formats displacing traditional modes. To download the document, open it in large view. The download button will appear in the top-right corner. #GenAI #AI #AIethics #TechnologicalAdvancements