Ralph Bolton’s Post

X’s New AI Settings: What You Need to Know and How to Opt-Out https://buff.ly/3y8DiQB via ChandraKumar R Pillai on Thinkers360 #AI #Analytics #Blockchain

Deep Fakes is clearly hard challenge to solve for any organization trying to leverage on the GenAI wave…as well as for us, media and content consumers trying to recognize what is truth..Big Tech is now trying to put in place measures to better detect AI-generated content (called C2PA which stands for Coalition for Content Provenance and Authenticity). This is certainly positive initiative..trying to include watermarks into the images and its metadata generated by BigTech LLM model. Now, this can surely be circumvented not only by taking screenshots as mentioned in the article but also by deep faking the watermarks themselves. Also, how we, poor consumers will identify these watermarks in all the content that we consume every day? I believe this is actually great opportunity for blending GenAI with Blockchain capabilities…imagine content curator (offered by your favourite brand virtual assistant) that is able to tell you and filter out for you only the content that has authorship recorded in immutable blockchain based platform network. And this will be the only content you allow yourself to see and read… thrilled to see who will be the first to deliver such magic coach #genai #artificialintelligence #deepfake

Has generative AI proved to be anything more than a glorified autocorrect? If startups are not building AI wrappers, they’re bolting AI to their product in order to increase their valuation, just as they did 5 years ago when blockchain was meant to be the next big thing, and VCs are eating it up. A cynical, but also very insightful blog post on the current state of generative AI. 'These things are not "learning," or "understanding," or even "intelligent" — they're giant math machines that, while impressive at first, can never assail the limits of a technology that doesn't actually know anything' https://lnkd.in/eW2XWfsQ

In the midst of the #AI revolution, #trust is the new battleground. 🌐💔 A new Edelman poll reveals global trust in AI has dipped from 61% in 2019 to 53%, highlighting a critical trust deficit. - Only 35% of US respondents trust AI, down from 50%. - The world views AI innovation as "badly managed" by a 2:1 margin. - Despite high trust in tech overall, AI faces skepticism. 𝐈𝐌𝐏𝐀𝐂𝐓: We're seeing the same pattern emerge that happened with #blockchain. As #marketers, it's critical to look beyond AI's mechanics to its true potential. Anecdotal conversations with #brands also reveal that while they're blown away by the tech, the impact may be overblown. Companies must focus on the "why" & "for whom" to rebuild trust. 🛠️🔍 #ResponsibleAI #TechEthics https://lnkd.in/eKgDis4i

in the generative Rise of ai, data is the new oil. So why shouldn't you be able to sell your own?.... #AI #Blockchain #Data #generativeAI #plans #Reddit #Rent #train #Users #Vana

Did you hear? Open AI is teaming up with Microsoft to launch a $2 million societal resilience fund to help drive the adoption and understanding of provenance standards Read more: https://lnkd.in/gE5bhuVx. DataTrails' platform adds provenance to metadata for AI solutions. #Provenance #AItransparency #betterthanblockchain

Salesforce's MINT-1T, a trillion-token dataset, revolutionizes AI research by democratizing access and raising crucial ethical questions about the future of machine learning. #tech #ai #blockchain #crypto #data #machinelearning #computing

The history of Blockchain is the future of GenAI. Molly White's Molly White's post on AI  in her great newsletter is very useful reading: "When I boil it down, I find my feelings about AI are actually pretty similar to my feelings about blockchains: they do a poor job of much of what people try to do with them, they can't do the things their creators claim they one day might, and many of the things they are well suited to do may not be altogether that beneficial. And while I do think that AI tools are more broadly useful than blockchains, they also come with similarly monstrous costs." https://lnkd.in/eNj-KFCF

This was a response to a post raised by Emmanuel. (I wanted to see if someone would go ballistic and try to 'pile drive' me. 😈 ) The point is that there's a deeper issue that isn't being discussed. Its the current infrastructure found in the enterprise. Its a mess. Because spending money on this is a zero sum gain. (No net new product or value stream...) You're spending money to improve your operational efficiency but you really can't show a strong ROI. You're going to spend more money that might reduce your operational costs down the road. (Which is ironically one of the arguments for Cloud that never really panned out... ) So how do we start this discussion that everyone knows we need to have, but no one wants to talk about it? -=- Emmanuel Maggiori, but can we really talk about AI? I mean its what everyone else is talking about? 🤯 🤬 Just kidding. Look, AI is the last mile. You want to focus on AI... make sure you can get to that last mile. Enterprises can't. Too many snake oil sales critters promoting this or that technology. (And its not just the sales critters... how many here think ELT is real? ) So lets talk about AI. But not focus on it but how to get your enterprise ready to do more than just experiment? That means a focus on infrastructure. Data and data governance/quality issues. That's the tricky part. And this problem will always be there.

At the very minimum, every enterprise needs to pose these scoping questions to teams asking for any #AIproject funding. What business problem is the AI solution trying to solve? What alternative solutions (rule based, advanced data & analytics) were considered for solving the business problem? Why is AI the only and best solution for this business use case? What are the cost-time-productivity savings (direct/indirect) for adopting AI? What are the estimated E-to-E costs for each AI model lifecycle phase (including post deployment monitoring, retraining, #RLHF, #MLSecOps)? Which audiences are impacted by solution predictive outputs and how should risk be measured for rights and safety impacts? Thanks to Emmanuel Maggiori for addressing the industry albatross/white elephant.