AI Scares the Crap out of Me

AI Scares the Crap out of Me

What do these three people have in common?

First, I think all of these individuals look very approachable and I wouldn't have any challenge with the idea that these individuals likely share some of the same values that I share and would be happy to have a conversation with them on social media or wherever simply based on their appearance. To me they look like me. I haven't met them, but I would tend to think they sound like me and even have similar values as me. However each of these images were created by AI. You can find the site where all of these individuals are at at: https://thispersondoesnotexist.com/.

Lately, I've started to look at Twitter profiles and noticed more and more that many of the people who are posting, especially relating to political messages seem to be bots. They are not, in many if not most cases, curated content by a social media manager or an individual who is trying to share their sincere political view. They are bots and likely early versions of AI which have a mission of swaying political opinion. The photos behind many of these "bots" look like real people. Individuals who had to have agreed to have their photos with their MAGA hats on in order to espouse a view. I only mention MAGA hats because they are definitely a tool of influence by the Trump campaign for better or for worse. Trump himself is a master NLP practitioner and as a result he knows how to create images which override rational thought by creating an image or creating cognitive dissonance such that his ideas can get passed through as an equally valid thought that confuses the masses. No one in the history of time has been as good as Trump has been. That being said, Synthetic Humans may soon have both the intelligence and the bandwidth to create the illusion of being a real person such that eventually it will be impossible to tell the difference. Think about the an army of Twitter AI bots that literally match the following, posting and engaging behavior of those who are thought leaders as real people. Right now bots are little bit easier to recognize, especially the bad ones, like the one I found the other day that has a nice woman's picture and a bio that reads God Country but has literally tweeted 375,000 times in the last few years. That being said that profile that I reference has 145,000 followers, so for better or for worse that profile is swaying public opinion. Put a true AI on it and the ability to create 10,000,000 other Twitter accounts over time and Twitter will be dead ultimately because no one will trust what they read on Twitter. I actually think this is fundamentally why Trump wasn't a big fan of the purge of bots by Twitter since he himself understands the power of the bot to spread a message. I myself read many tweets with a jaded perspective and question is this real or is this a bot... I don't have an issue with good quality curated content but at some point the bots become the message and then the platform has no value any more. I don't think we are far from, if it hasn't happened already, where we start quoting AI bots as if they were humans because we don't know how to filter them out and as a result the bots will get published in an article pointing out a position thinking the bot is a human taking a position.

Recently I attend the Abundance 360 event in Los Angeles and on display were a couple instances of Synthetic Humans. The most high profile was from a company called Soul Machines. https://www.soulmachines.com/. This AI was just enough not real to not overly complicate the AI ethical issues. It was good, and maybe as good as todays technology would allow for Synthetic Human interaction however it isn't far off.

I was also thinking this last week about content on the web. Right now most content is still humanly curated. However earlier this week a team that had created an AI to write research papers actually decided not to release it because the content was too convincing that it was writing well researched articles. https://www.cnn.com/2019/02/18/tech/dangerous-ai-text-generator/index.html

About two years ago I got a phone call from a charity. There were a couple oddities on the call but not enough for me to catch while I was on the call. I even committed to putting $50 toward the charity because of the quality of the conversation however I had the uneasy feeling that I had just talked to an AI. Fast forward to today and watch this video from CNET on Google Duplex.


Now we add in the idea of working virtually with a virtual human who seems human and eventually we truly reach Ray Kurzweil's Singularity.

My big challenge isn't that certain aspects of AI won't help us. I look at my G-Suite, Alexa and even Siri on my phone and those aspects of AI really do add value, however when AI's start to shape our thinking without our knowledge and without our consent and they skip all the normal ways that we would filter real people and drill straight into our decision making abilities it gets pretty scary pretty quickly. Free will is shaped by the opinions of others however what happens when free will is shaped by the opinions of a very few who ultimately have an army of collaborating AI's all designed to recontexualize the world in favor of any view? With enough data points from rapidly created AI research reports and web pages that are created that reinforce those views, along with AI's which post and engage like humans on social media pushing those views, to the phone calls that one will receive from synthetic humans who you would have no idea are not real humans but know better than you how you make decisions and can shape your view to virtual humans who are on video, have photos of themselves taken in real settings that are total illusions it will not be long before we cocoon for safety purposes. Faith Popcorn in the early 80's talked about cocooning and I expect there will be a purge of social media friends and who you follow not because in the early days those social media friends and following of others made sense however in the not to distant future we may start to not trust anything that doesn't come from someone we know and trust.


Dana DiPasquale

Sales & Business Development Executive: Customer Success, Portfolio & Pipeline Management, Human-centered Solutions, Contract Negotiation, Realtech/Proptech

1mo

I fear we might say in the future "gone are the days for human-centered solutions".... Intuition and AI can complement each other however, AI domination scares the crap out of me too.

Like
Reply

I have long-believed the Singularity would offer in my lifetime. It seems a lot closer now.

Like
Reply
Monikaben Lala

Chief Marketing Officer | Product MVP Expert | Lead Gen Specialist

1y

Glenn, thanks for sharing!

Like
Reply
Nada Djomehri

Using real estate as vehicle steering vision to reality

4y

Thanks for stirring our mind and heart-set, Glen! What rings loud for me here is the age-old issue of insincere intent to mislead ideas and steer the masses (probably cause it's easier). While the probability of engineers developing instruments to differentiate synthetic from natural would be likely in this rapid wave of opportunistic technological advancements. I still stand to believe that "closed mouths only add to the oppression on impression" and that to filter, confront and hold culprits accountable (although it may be challenging), is collectively far-better, than resorting to band-aid fixes. #humanity #AI #ethicalhacking 

Like
Reply
Corey Engel 🧢

👉 Building software/companies | family office/private equity backed | powered by Family & relentless gratitude | (Amazing wife and 5 kids)

4y

Glenn Sanford there’s 10X amount of opportunity to engage agents and leverage the data gathered from top producers to train inbound agents. AI is so misunderstood it’s crazy. We’ve personally built it for insurance, healthcare, banking, etc. and something interesting is happening. Not only are we capturing 20-30% of lost revenue, AND creating an environment where contractors actually WANT to use company tech, we also are helping bring back humanity to companies because users enjoy their jobs, get more done, are more productive, and spend newly found time on family and better company culture.

To view or add a comment, sign in

Insights from the community

Others also viewed

Explore topics