Kurt Smith

My BS meter went off recently. Key lesson in web3... When it sounds amazing and I'm excited. 🛑 This is when I STOP. The digital colonoscopy STARTS. – Google and do tons of research – Ask for references and talk to them – I'm patient with any signed agreements There's absolutely no need to rush into something important. Wishing you all a great Friday! 🚀 Follow me to learn all about web3 ---- Why listen to me? – 3,500+ hours experience running projects in web3 – Scaled my project to 70,000 holders in 3 weeks – Made every mistake in the book – Learned from near all of them – Follow me to avoid them #web3 #blockchain #innovation #startups #technology

49

29 Comments

The Man Who Predicted AGI Decades Ago w/ Ray Kurzweil | EP #125 Author: Peter H. Diamandis Ray Kerswell and Peter Diamandis have a wide-ranging conversation about the future of AI, AGI, longevity, and more. They discuss Ray's predictions for AGI and the singularity, potential job loss in the future, advancements in nanotechnology, the implications of ASI and AI merging with humans, and the future role of AI in centralization or decentralization. They also address the impact of AI on medicine and the potential for reaching longevity escape velocity. Ray believes that AI will democratize capabilities, enabling individuals to have more impact and potentially passing the Turing test in the near future. The conversation ends on a positive note, highlighting the extraordinary time to be alive and the potential for an even more spectacular future. https://lnkd.in/eqx4__7Q ___ Generated by Smart GPT

1

This is an interesting question. If an AI chatbot presents false, potentially libelous information, can it be prosecuted, and according to what laws? There are clearly examples of hallucinations from Google. However, as the article points out, proving intent on the part of the AI chatbot, as well as its creators, could prove somewhat difficult. What do you think? I'd say the question of value vs. risk within the context of generative AI is getting more complicated. It seems clear that some use cases for AI provide quantifiable value and low risk. Others less so. I suspect the AI-craze will die down a bit as companies increasingly focus on lower risk use cases that can be confidently linked to an ROI. #genai #llms #chatgpt #openai #libel #hallucinations Read more: https://lnkd.in/gz994JTy

4

Hot tip for an aspiring startup founder or DefTech VC: find an incumbent vendor with large, sole-source fixed-price contracts, and make whatever that is… Had a convo with a govy customer yesterday, being the sole source for a product while also being on a FFP contract enables you to basically charge whatever you want, at whatever margins you want. If your thing is important, and the government often buys it, you effectively get monopoly power, and because you’re of FFP contracts, they can’t audit your margins. Now here’s the opportunity: companies who over play/press this market advantage tend to make adversaries out of their customers, and those customers often would gladly diversify their sources of supply if they could… Now, this mostly works with big end items like bombs, missiles, and similar expendables, but I’m sure a little time on SAM.gov and you’ll find some market asymmetries.

66

19 Comments

In May 2020, a ripple began in the vast ocean of AI and copyright law—a ripple that has since turned into a wave poised to reshape the information ecosystem as we know it. 📜 Thomson Reuters filed a pivotal lawsuit against Ross Intelligence, positioning itself against the nascent AI industry. This was just the start. The spotlight today shines brighter with every proceeding taking place across U.S. courts. But here's the kicker: What started as a niche legal battle has escalated into a nationwide showdown, drawing in unthinkable stakeholders—from authors like Sarah Silverman to giants like Universal Music Group, and tech behemoths like OpenAI and Meta. Central to this legal maelstrom is the doctrine of “fair use.” At its core, this principle is sparking debates—should training AI on copyrighted material without consent be deemed fair? The stakes are high, and the outcomes could redefine creativity and technology. WIRED’s latest visualizations make it easier to track these unfolding cases, showing who’s up against whom and the arguments in the spotlight. This legal saga isn’t just about AI companies versus copyright holders. It’s about defining the boundaries of creativity, innovation, and ownership in a digital age. What’s your take on AI’s role in the creative landscape? Is it a partner in innovation or a challenger to traditional rights? #AICopyright #FairUse #InnovationInJeopardy #TechPolicy

1

Kudos to Mignon Fogarty for her post in response to NaNoWriMo's statement on AI, and for resigning from their writers' board. If you're curious about the controversy, I encourage you to check the link. National Novel Writing Month (NaNoWriMo)'s AI statement was released just a few days ago, and it's already on at least its third iteration. Generative AI is a polarizing subject, especially among fiction writers. NaNoWriMo's statement of position to the fiction writers who keep it going was Not Good. #NaNoWriMo #WritingAI #FictionAI #NaNoWriMoAI

43

4 Comments

A math problem straight out of a Danny Boyle film... $250M in pandemic aid fraud with $50M recovered leaves at least $200M. There's probably a lot more that hasn't been identified yet (no one stops after their first $250M), but we'll call it $200M. Getting a jury to lean your way doesn't take all 18 jurors, but even if you paid them all off including post-trial bonuses you're up to something like $3M in costs. Spending 1.5% of what you stand to lose seems like a solid business case. It would be strange to find that kind of strategy unbelievable if you've worked around fraud much, but this is the quote used in the The Associated Press article that was picked up by a lot of news sites: “𝑇ℎ𝑖𝑠 𝑖𝑠 𝑐𝑜𝑚𝑝𝑙𝑒𝑡𝑒𝑙𝑦 𝑏𝑒𝑦𝑜𝑛𝑑 𝑡ℎ𝑒 𝑝𝑎𝑙𝑒,” 𝐴𝑠𝑠𝑖𝑠𝑡𝑎𝑛𝑡 𝑈.𝑆. 𝐴𝑡𝑡𝑜𝑟𝑛𝑒𝑦 𝐽𝑜𝑠𝑒𝑝ℎ 𝑇ℎ𝑜𝑚𝑝𝑠𝑜𝑛 𝑠𝑎𝑖𝑑 𝑖𝑛 𝑐𝑜𝑢𝑟𝑡 𝑜𝑛 𝑀𝑜𝑛𝑑𝑎𝑦. “𝑇ℎ𝑖𝑠 𝑖𝑠 𝑜𝑢𝑡𝑟𝑎𝑔𝑒𝑜𝑢𝑠 𝑏𝑒ℎ𝑎𝑣𝑖𝑜𝑟. 𝑇ℎ𝑖𝑠 𝑖𝑠 𝑠𝑡𝑢𝑓𝑓 𝑡ℎ𝑎𝑡 ℎ𝑎𝑝𝑝𝑒𝑛𝑠 𝑖𝑛 𝑚𝑜𝑏 𝑚𝑜𝑣𝑖𝑒𝑠.” Maybe Thompson is trying to color public opinion by drawing a link between Goodfellas and the defendants. If so, he forgot about the fact that everyone loves Goodfellas. Fraud trial juror reports getting bag of $120,000 and promise of more if she’ll acquit (https://lnkd.in/eYRGWFWa) #fraud #fraudarticle #fraudcharges #aidfraud #charity #law #jury #crime #organizedcrime #movies #bagofcash

7

The U.S. Department of Justice (DOJ) is pushing for a major antitrust action against Google, aiming to dismantle its monopoly over search and digital advertising markets. The DOJ claims that Google’s dominance is built on exclusive agreements that make its services the default on devices and browsers, limiting competition. A recent court ruling supported the DOJ’s case, and structural remedies are being considered, potentially breaking up key services like Chrome, Android, and the Google Play Store to restore fair competition. Google argues these measures are “radical” and could harm American innovation, but the DOJ’s refined plan will be revealed later this year, possibly setting a regulatory precedent for Big Tech. #DOJ #Google #Antitrust #Monopoly #BigTech #TechRegulation #Innovation #CompetitionLaw

3

AI's massive theft of Intellectual Property with no attribution to copyright is a fundamental flaw in the black box AI architecture. Only major players such as the New York Times can sue. Class actions are hampered by cost, slowness to get justice, and difficulty to pull evidence from a black box. Tragic that AI makes the open Internet dumb. Anything of any value will be placed behind paywalls. The incredible STUPIDITY of AI architecture manifest with updates forcing multi-billion dollars rebuilds of Large Language Models. Imagine if databases would have to rebuilt every time a record is added. The waste of energy to rebuild and operate the LLMs is becoming astronomic requiring gas and nuclear power to be built with consumers footing the bill with skyrocketing energy prices and horrific consequences to the environment. Forget about net zero 2050. #openai #microsoft #google #llm #architecture #copyright #ip

4

7 Comments

Who is reading all those meeting transcripts? I feel like we need an AI assistant to summarize all of the meeting transcripts for the day or week. I know this is coming. Don't get me wrong, AI transcriptions help with capturing decisions, action items and meeting summaries but reading the transcript is not likely. It will be interesting to see where meeting go in the future. All AI assistants? Southern California SIM (SCSIM) Society for Information Management #AI #CIO #toomanymaatings

16

11 Comments

Paranoia 101: There ARE times when what is presented to you is simply that. No ulterior motives, no hidden agendas, a distinct lack of Illuminati, government, aliens, or hollow earth folks. Take for example this exchange: Hey, do you know of any good 3rd party remote access tools? (Me) Yep, several, what are you looking to do Just to be able to log in to my home machine from anywhere (Me) Here's a few, take a look, I think I have a license for one of them That'd be great, thanks! (Me) Most welcome ....two days later Um, so, that remote software was great (Me) Splendid, glad it worked! But, I was talking with a friend, they said you're a hacker (Me) Yep And that you do things for the government (Me) Possibly And that you've probably given me software that can be remotely accessed by you and anyone you want at anytime, and that you might be spying on me, or copying/taking my data. In fact they think you might have installed a trojan or something to listen into my conversations... (Me) Was this friend wearing a tin foil hat? (enquiring minds etc.) Um, nope, but they are concerned... (Me) Fair, very fair. What I gave you was a legitimate license for a piece of software that would allow you (and only you if you followed the security advice) to access your system from wherever you needed/wanted. But could you get into my machine? (Me) Yep, but not using that software, I'd simply go in through another way, I mean the operating system you are using is Swiss cheese, the ISP sells your data every day, the clouds and social media systems you traverse hourly tell me what you're doing, where you're at, and how often you're in the bathroom, and you use the same set of passwords for almost everything. I don't HAVE to break in when you hand it over on a silver platter. Ah, um, hang on, how do you know about my passwords? (Me) We're humans, most follow similar patterns. Oh... Hey, can you teach me some of this? (Me) Yep, just don't EVER ask me to fix your printer, deal? 'all for now Chris #technology #geek #infosec #Cybersecurity #hacker #conspiracy #givernment #help #community #awareness

217

82 Comments

1 in 4 IT leaders regret investing in GenAI so early. The keyword there is "early." So many companies over the last 24 months have adopted for the sake of adoption. The risk we run now is that the technology is being blamed for the outcome, not the poor selection of use cases. Let's be clear: AI is not going anywhere and you will need it to be successful in the long run. It may not be obvious now, but just like the web, it will become ubiquitous and mandatory. Don't let the early false starts fool you. https://lnkd.in/gJ3HB_VU

39

8 Comments

50% of federal agencies already have a Chief AI Officer (CAIO). Why? Because AI is changing everything. Think cybersecurity, efficiency, and innovation - all powered by AI. But here's the catch: most agencies are still figuring out how to use AI effectively. That's where a CAIO comes in. They: * Build a skilled AI workforce * Ensure AI is used ethically and transparently * Guide agencies in adopting cutting-edge AI solutions For executive recruiters, this is a goldmine. The demand for CAIOs is exploding, and finding the right leaders with the right skills is critical. P.S. - Do you think we will see more AI officers in the recent future?

6

5 Comments

BENEATH THE SURFACE | Plenty of hot takes and rightful outrage over the OpenAI debacle involving Scarlett Johansson's likeness and "Her" (movie pun intended) voice in GPT-4o. If you need a rant, look no further than posts from your favorite LinkedIn Top Voice AI InfLUenCerS. The only thing I'll add to that noise is it wasn't due to stupidity or oversight; it was driven by utter hubris. Moving on, the more insightful discussion revolves around the potential legal action now that Johansson has hired counsel. Look, this nation—nudge elected representatives here—is overdue for serious antimonopoly/antitrust measures, informed consent considerations, and baseline protections for intellectual property. Those actions require commitment, courage, and informed nuance. Unlike that time when Congress asked Google's CEO about Apple's iPhone, or when they were confused whether TikTok could connect to WiFi. It's high time for precedents, both within and outside the court. Such precedent has been wholly lacking. Section 230 remains a mess, SCOTUS has splintered and/or punted big tech cases, and meanwhile, the states' patchwork of legal interpretations looks like your childhood blankey that's seen better days. But precedent—that’s what this OpenAI situation has opened the door for. Now, let’s look to the past. Remember the McDonald's hot coffee case from 1992, when 79-year-old Stella Liebeck spilled coffee on her lap and a jury awarded nearly $3 million in punitive damages for the burns? The common reaction is, "Wait, how? Isn’t coffee supposed to be hot?!" Yes, but it was dangerously hot and the courts had to compel the company not to be negligent again. For you or me, or a mom-and-pop shop, that might mean a $20k medical bill we're liable for. But for McDonald's, it needed to be several days of coffee revenue plus the medical bills. Voila! You now better understand the concept of punitive (or exemplary) damages. So, what's the obvious question regarding OpenAI? It is, "What would it take to sufficiently compel a company valued at ~$80 billion to not operate with callous disregard for intellectual property and consumer protections?" Additionally, "What precedent would compel others, too?" Side note: It’s interesting that OpenAI decided to remove Sky's voice, essentially validating and adding material relevance to the claim. That’s the call-to-action and stakes beyond the hot takes. Is there a case? Yes! Look no further than Bette Midler v. Ford Motor Co. (1988), where the courts ruled that the voice of someone famous is distinctive to their person and image. Therefore, it is unlawful to imitate their voice without express consent and approval. The ruling awarded Midler $400k for a single commercial, equivalent to ~$1.1 million today. This is not about correcting a mistake—it’s about precedent to protect intellectual property and consumer rights on a larger scale in the age of generative AI. #technology #artificialintelligence #openai #law

40

10 Comments