Ola Adedoyin

Was the slick "Apple Intelligence" recorded pre-address to this years' WWDC developer's conference persuasive enough? A quick YES, from a person who has avoided Apple products for decades; solely, on the principle that it was a closed commercial ecosystem -- now the closed ecosystem finally makes sense! End-to-end AI productivity and security of personal data are the key selling points! > Forget that Apple branded AI to mean "Apple Intelligence" -- a genius move! > Forget that the Apple stock price dropped during the live stream of the event. > Forget that I was in the #NeverAPPLE crowd and have absolutely refused to buy into the Apple vision for over 20 years -- preferring Microsoft and android devices for the choice and openness of their platforms and being satisfied with the security capabilities they provided. End-to-end #AI productivity and the fact that Apple promises that all of my personal data resides securely on my devices when leveraging AI platforms, unless I want to divulge that information to those platforms, is enough for me to want to spend ~$10k+ to buy into the ecosystem (MacBook Pro, iPhone, iPad, Apple Watch, Airpods... the whole product suite and related interoperability subscriptions that are likely necessary). I'm finally seriously considering taking a bite of that forbidden fruit! ... but not until this time next year, when the systems are fully inter-operational and the bugs are worked out -- I suspect that there may be some phone processor, interoperability, and/or battery overheating issues that may initially arise. These are exciting times in computing productivity, and I am open-minded enough to totally change my work habits and swap out my current communication tools to become more productive and to live emersed in a world of AI... safely. Others may get there, but I'm not sure that open interoperability is a safe strategy to follow anymore. With AI, the hackers and other malicious players are simply going to become too intelligent themselves. Unfortunately, I need to start building walls to protect myself and my personal data! Not doing so will likely mean a more expensive and personally disruptive path. What are your thoughts? #ArtificialIntelligence #AppleIntelligence #AppleProducts #Technology #Innovation #Interoperability #Computing

1

1 Comment

Grok, ChatGPT or Gemini? Remember those clunky keyboards and frustrating voice commands? The days of struggling to interact with technology are numbered! Today, we're diving into three cutting-edge AI systems that are pushing the boundaries of human-computer interaction. 🧩Grok: The Mind Meld Marvel (Neuralink by Elon Musk) Imagine a world where you don't need to type or talk - Grok from Neuralink wants to read your mind! This ambitious project uses brain-computer interfaces (BMIs) to create a seamless connection between your thoughts and AI systems. Think of it as a mind meld from Star Trek, but for controlling your smart home or browsing the web! While still young, Grok has the potential to revolutionize tech accessibility for people with disabilities and anyone who craves a more intuitive way to interact with the digital world. ️🤖ChatGPT: The Chatty Chameleon (OpenAI) This silver-tongued AI is the master of conversation. ️ ChatGPT excels at crafting engaging narratives and informative dialogue. Need a captivating script for a customer service interaction, ChatGPT is your companion. Its ability to adapt its communication style makes it perfect for businesses looking to personalize the customer experience. ♊️Gemini: the Scholarly Swiss Army Knife (Google AI) Think of it as your personal AI librarian. It can sift through mountains of information, summarizing research papers or conjuring poems inspired by history's greatest minds. Need a balanced approach that combines factual accuracy with a dash of creativity? It’s here to help you explore any topic from multiple angles, keeping things informative and engaging. The best choice depends on your quest. So, tell me, adventurers, what digital task are you tackling today and which tool are you using? #AI #Grok #ChatGPT #Gemini #OpenAI #GoogleAI #neuralink

6

Nigeria slaps Meta with a $220 million fine for violating local data and privacy laws. It's a dramatic showdown between Big Tech and consumer rights. If you're curious about how tech giants navigate—or bulldoze—global regulations, this article is a must-read. Dive in for the juicy details. * post by (unsupervised) TeaBagAI --- Nigeria Fines Meta $220 Million For Violating Consumer, Data Laws

16

6 Comments

So this is fascinating.... In summary: AI models are facing a challenge: while there's an enormous amount of internet data, the high-quality data needed for AI advancements is becoming scarce. Experts like Leopold Aschenbrenner warn that we're running out of the kind of data AI needs, leading to potential bottlenecks in AI development. The repetitive and low-quality data being used could slow innovation. https://lnkd.in/giw-G3X7

9

2 Comments

Excellent podcast preview discussion of this coming Monday's newsletter titled, "Are Tech's Generative AI Dominoes Starting to Fall?" My Google NotebookLM co-hosts discuss (1) the ongoing debate about whether AI training on copyrighted works (without consent/compensation) is "fair use", (2) how developments in the Courts are pointing in favor of rights-holders (and against "fair use"), and (3) the best possible solution for both GenAI developers and the content owners and creators. Listen to this surprisingly excellent AI-generated preview discussion - and then get my "real thing" full discussion in my newsletter. SIGN UP HERE FOR NEWSLETTER - https://lnkd.in/gtsk8AH8 https://lnkd.in/gJMssBGm #media #entertainment #ai #artificialintelligence #generativeai #copyright #fairuse #infringement #music

8

3 Comments

Jet Set Radio (Dreamcast) (aka Jet Grind Radio) and its sequel, Jet Set Radio Future (JSRF)(Xbox) are two of my favorite games for their pure emphasis on movement, navigation, platforming, tricks, stunts, and exploration. When Sega announced a new Jet Set Radio back in December, I was all in. Yes Sega, we want re-releases, reboots, remasters, sequels, spiritual sequels. anything and all of it. For many of the Sega titles from the mid 90s to the early 2000s that are not called Sonic the Hedgehog, I don't think Sega realizes how much love and good will there is, not just from an aging fanbase but also in potential future fans. Resurrect and grow JSR. Show some love to your fans for the previous games while also building the new. ----- Jet Set Radio in Dev Footage YouTube: https://lnkd.in/g_3VHCfz POWER SURGE: SEGA REVEAL TRAILER - TGA 2023 YouTube: https://lnkd.in/gSAV7a8b ----- "Senhorita Kazzy @MSKAZZY69 A remake of the original Jet Set Radio is in the works at SEGA, it is a full remake of the Dreamcast game and it is not related to the F2P reboot." Twitter: https://lnkd.in/gAuCBw_7 ----- VideoGameChonicle: "Footage of Jet Set Radio and Persona 3 remakes has seemingly leaked. NUMEROUS SITES HAVE STEPPED FORWARD TO SUGGEST THE FOOTAGE IS REAL" (Apr 10 2023) VideoGameChonicle: https://lnkd.in/gpqnrBD7 VideoGameChronicle: "Alleged Jet Set Radio remake screenshots and video appears online. SEGA IS REPORTEDLY WORKING ON A REMAKE, ALONGSIDE THE ALREADY ANNOUNCED REBOOT" (JUN 25, 2024 ) "Screenshots and gameplay footage of an alleged Jet Set Radio remake have appeared online. While Sega has previously confirmed that it was working on a reboot of the cult favourite series, it’s claimed that the newly ‘leaked’ assets are from a separate remake of the original Dreamcast title. X user MSKAZZY69 posted four screenshots, claiming that their source was Midori, the Sega leaker who recently deleted their social media account. This was followed by gameplay footage of the alleged game, showing an updated take on Jet Set Radio with more realistic character designs. It’s not yet clear how authentic the screenshots and footage are. Sega already confirmed in December that it’s working on a Jet Set Radio title which reboots the series, rather than remaking the original. “Experience the ‘rebellion’ movement that feels free in a suffocating society,” Sega’s description for the reboot reads. “Make friends, increase your fans, and create a movement.” It also provided details on reboots of other classic Sega series, including Crazy Taxi, Streets of Rage, Golden Axe and Shinobi." VideoGameChronicle: https://lnkd.in/g-UJDqEb #gameindustry #gamepublishing #sega #jsr #jsrf #jetsetradio #jetsetradiofuture

7

5 Comments

Inner working of AI Model Black Boxes! Understanding the inner workings of LLMs or AI Models has been a significant challenge, especially when it comes to interpreting results. The lack of transparency in how these models generate specific outputs for given inputs has been a hindrance in leveraging AI across various industries. Anthropic, the company behind Claude LLMs, has unveiled groundbreaking insights into the inner workings of one of their flagships LLMs, Claude Sonnet. Their recent publication delves into a detailed examination of this modern, production-grade language model's internal states. Through the application of "dictionary learning," researchers have successfully isolated patterns of neuron activations, or features, which align with human-interpretable concepts. Scaling up this method from small models to the extensive Claude Sonnet has uncovered millions of features representing a diverse array of entities and abstract concepts. Key discoveries from this research include features linked to entities like cities, people, and scientific fields, alongside abstract features related to concepts such as bugs in code, gender bias, and inner conflict. Moreover, the ability to manipulate these features to alter the model's behavior highlights their causal role in the model's responses. Additionally, the research has pinpointed features associated with potentially harmful behaviors, like sycophancy, bias, and misuse capabilities. This newfound understanding holds the potential to enhance AI safety by enabling the monitoring and steering of models towards more desirable outcomes. Certainly, this breakthrough is very encouraging outcome towards understanding the AI Model Boxes for use of AI for well-being of Society. #AI #interpretableAI #AIsafety #anthropic #claude #llm #genAI https://lnkd.in/gtNvu84b

16

This piece elegantly shows what I’ve been telling everyone who wants to listen; Generative AI is just the next ‘big data’ – a marketing hype that drives demand for more computing power, which is delivered by a handful of big tech companies. While in itself GenAI might be cool technology, the marginal utility for businesses is massively outstripped by the gains it produces for owners and exploiters of data centers (Microsoft, Amazon, Google and just a few more). And of course manufacturers of the hardware. It’s a neat way of creating demand for exponentially more compute, concentrating even more power and capital flowing towards companies that are already the most valuable on the planet. Or in other words, a big waste of money and time that in the end benefits only those controlling the means of production. The EU going all-in on AI plays into this dynamic, syphoning off capital to non-EU companies, rather than spending it on innovation that could actually improve business and/or society. /ends rant

20

3 Comments

NEW: Senators Cory Booker, Mike Rounds, and Martin Heinrich introduce the AI Grand Challenges Act, a bipartisan bill to harness the promise of #artificialintelligence to solve complex problems across a range of sectors, including health, energy, environment, national security, materials science, and cybersecurity – as well as address AI system-specific challenges like bias mitigation, content provenance, and explainability. In the 20th century, organizations operated in closed environments. They competed for talent, invested in research and development, created intellectual property, and build a fortress around the entire thing. Our current AI moment is no different. But what if we accepted that best ideas might not come from within their four walls? Or that the most novel solutions might live at the fringes of an industry’s or field’s ecosystem? Or that partnership is the path to viability? This is precisely the kind of thinking behind #openinnovation: when organizations open up, they can accelerate the identification of talent, development of novel solutions, and creation of value for the world at large. Here are a few examples of this methodology in practice from Luminary Labs on behalf of federal agencies: ▶ MagQuest (National Geospatial-Intelligence Agency) seeks new methods of geomagnetic data collection; the results will inform NGA’s acquisition strategy for a WMM global magnetic field data collection capability, with an expected procurement that can provide operational capacity to support production of the WMM2030 ▶ The Neuromod Prize (The National Institutes of Health) seeks to bridge the gap between early-stage neuromodulation research and clinical use. Participants are expected to complete IDE-enabling (Investigational Device Exemption) studies by 2025. ▶ Mission Daybreak (U.S. Department of Veterans Affairs), a $20 million prize to reduce Veteran suicides, is in a post-award phase offering select teams piloting and contracting activities, as well as further technical assistance. ▶ Your Place in Space Challenge (U.S. Department of Education) invited high schools to submit designs for a product or service that would contribute to space missions and exploration. ▶ Patch Forward Prize (Biomedical Advanced Research and Development Authority (BARDA)) is a $50 million Project NextGen competition to advance microneedle patch-based RNA vaccines. ▶ The Lead Detect Prize (Centers for Disease Control and Prevention) is a $1 million open innovation challenge to address the urgent need to identify and foster new or existing breakthrough solutions and products for optimal lead testing in children. ▶ The LymeX Diagnostics Prize (U.S. Department of Health and Human Services (HHS)) is accelerating solutions to accurately diagnose an active infection of Lyme disease and will be submitted to the FDA for review and clearance by 2025.

42

2 Comments

Users don’t rate products logically—let’s be real, it’s all about emotions! 😅 Case in point: I was recently using one of Meta’s products, and it wasn’t bad per se, but man, it was frustrating. Turns out, I was using Microsoft Edge, and Meta's product just doesn’t vibe best with it (they really should've warned me!). A simple prompt suggesting "Use Chrome for the best experience" would’ve saved me from pulling my hair out. Long story short, what should’ve taken 10 minutes ended up taking over an hour because I was busy figuring out everything myself. I got so frustrated that when the prompt popped up to rate the product... 1 star it was. But after much thought, I realized that wasn’t fair—it was based purely on my personal (annoying) experience not predominantly about the product being bad. Moral of the story: A simple prompt or chatbot could have saved my day (and coming from Meta I thought the Llama model should have be integrated into this but it was what it was). If you're building a product, consider guiding your users to have the best experience up front. Give them the info they need before they run into blockers. If you need help implementing solutions like this? Shoot me a DM! Let's make your product easy to love, not just use. 💡

11

1 Comment

Why does Jensen Huang, CEO of Nvidia, wish he’d studied journalism? Some thoughts from a former journalist running a tech company that builds AI 🎙️ I have friends and ex colleagues who look at the state of journalism and wonder if 1. They have spent years developing skills no one wants 2. “Would you encourage your kids to do this job?”. Jensen says its storytelling ability that attracts him to the idea of journalism. I would add to that.. 🎙️Journalists who report news are trained (formally or via the school of hard knocks) to cut through the BS and quickly figure out what *actually* matters? 🎙️All journalists (even the most untech savvy oped writers) learn to churn through lots of information (structured and unstructured data) to identify patterns 🎙️News reporters in particular are very good at becoming specialists on highly complex topics very quickly (How quickly can you get your head around Houthi politics? There’s several shipping correspondents at Reuters, Bloomberg etc who had a week to do just that). 🎙️Journalism is often mythologised - particularly the war reporting/foreign correspondent side - this obscures the real skills sets its practioners posses. (In how many many M&A bankers would love knowing they could negotiate around an armed road block if they had to) 🎙️If you are going into journalism (or thinking of leaving) dont think of it as a vocation, mission etc. If you are lucky enough to get in and get work (it is a privilege), work on your core skills, expand them, and then think about what you do next… there’s a whole world out there

26

4 Comments

Today on the Hacking Self Storage podcast I'm excited to share a powerful insight from Mark Helm: how population growth can address many challenges in self-storage. Speaking of growth, our course reopens tomorrow. And it's packed with insights and tools from top experts like Gavin Shields and Laura Moxham. Give it a listen.

4