Matthias Hohensee’s Post

Nice development!!! This would be again earlier than everybody may have thought. does the quality of those models improve that fast then we can reconsider quite some considerations... While many I talk to expected it to be there maybe in four or five years that's the current offline models are so inferior, I have always been pretty confident that maybe in one and a half or two years from now we would have it. this development of Apple I think shivs that we may have offline models in comparable quality to what we see online today in half a year or a year. I think this is a game changer for many considerations we have for new products leveraging this technology. concerning data compliance especially gdpr this could be a big deal if we could just skip any off-device data processing for such products soon. "The 30B model achieves "competitive performance in all categories" compared to similar multimodal models, it says. With a model size of 30 billion parameters, MM1 would be smaller than the large language models of other leading companies: Meta's LLaMA-2 comes in at 70 billion, the Luminous model from German company Aleph Alpha is set to be expanded to 300 billion, and GPT4 is already thought to have more than a trillion parameters. Nevertheless, independent observers do not see the small model size of MM1 as a disadvantage, on the contrary." adiutaByte MEDIFOX DAN ResMed #innovation #GDPR #DSGVO https://lnkd.in/eATPrCaS

The countdown is on. Apple's global developer conference, the company's most important event, kicks off next Monday in Silicon Valley.  In his keynote speech, CEO Tim Cook must finally explain which AI strategy he intends to pursue and how this will be expressed in concrete products. Allegedly, a cooperation with OpenAI is to be announced. At first glance, this seems to be an odd couple, as OpenAI has been heavily criticized for its attitude towards privacy and data protection, while Cook has been selling this as Apple's core philosophy for years. However, OpenAI currently appears to have the most powerful generative AI - and Apple urgently needs to revamp its personal assistant Siri. Siri was introduced no less than 12 years ago and still can't hold a conversation. OpenAI also has close ties with competitor Microsoft. But its CEO Satya Nadella certainly has no problem with the Apple alliance if it benefits his Azure business.  It will also be interesting to see whether an alliance with OpenAI precludes another with Google or Anthropic. Unlike NVIDIA and Microsoft, Apple shares have not yet benefited from the so-called AI boost. However, Apple is well positioned thanks to its iPhone market power. Although it had to cede the global market leadership in smartphones to Samsung Electronics, the iPhone is still a bastion. Most consumers will use AI via their smartphone; neither Microsoft nor Amazon have their own smartphones. Google does, but the market share of its Pixel series is very small. If Cook manages the AI turnaround, this could lead to an upgrade cycle for iPhones and probably higher prices for the flagship model. Apple desperately needs a catalyst for its hardware business. AI could provide it if the services are actually useful. https://lnkd.in/eYn3bqpQ

🚀 The Next "iPhone Moment" with AI? 🤖 Pretty exciting times ahead as OpenAI's founder, Sam Altman, is reportedly working on a groundbreaking product for Artificial Intelligence. Remember the buzz when #ChatGPT showcased the potential of AI last November? Now, #OpenAI is in advanced talks with former Apple designer Sir Jony #Ive and #Softbank founder Masayoshi Son to create what could be the "iPhone of Artificial Intelligence." As someone working in IT I'm thrilled to see the convergence of design, technology, and finance to shape the future of AI. The collaboration aims to offer an intuitive user experience for AI interaction, reminiscent of the iPhone's impact. With Softbank potentially contributing over $1 billion and the expertise of Ive, who played a pivotal role in the iPhone's creation, the possibilities are endless. The tech world is abuzz with what's next after smartphones, and it seems AI might just be the answer. Looking forward to what OpenAI and its partners bring to the table! More in the article (in #German) 👇 #AI #OpenAI #Innovation #TechNews https://lnkd.in/ecmBVuyQ

We Should Make Our Own Decisions - AI Regulation in the EU In the rapidly evolving landscape of artificial intelligence, the European Union's stringent regulatory stance threatens to stifle innovation and competitiveness. The recently proposed EU AI Act, while aimed at ensuring ethical AI development, might inadvertently place European businesses at a significant disadvantage on the global stage. A recent Handelsblatt article highlights a glaring example: Apple's latest AI functionalities are expected to be delayed in the EU market until at least 2025 due to these regulatory hurdles. This delay underscores a broader issue where over-regulation hampers technological advancement and market responsiveness. The EU AI Act intends to establish comprehensive guidelines for AI usage, emphasizing risk management and ethical considerations. However, the stringent requirements could lead to a scenario where companies either divert their innovations to less regulated regions or face substantial compliance costs, thereby diminishing their competitive edge. As someone who values personal autonomy and innovation, I find this regulatory approach concerning. It raises a critical question: Are we, in our quest for ethical AI, inadvertently curbing our ability to innovate and compete? The answer seems to lean towards a resounding yes. Policymakers must recognize the fine balance between regulation and innovation. Instead of imposing rigid frameworks that stifle growth, we need policies that foster a conducive environment for technological advancement while safeguarding ethical standards. It’s not about discarding regulation but about finding a middle ground that doesn’t sacrifice our competitive edge. The global AI race is intensifying, and Europe cannot afford to lag behind due to overly cautious regulations. We must advocate for smarter, innovation-friendly policies that empower businesses to thrive while maintaining ethical AI standards. It’s time for a shift in perspective, and our message to policymakers is clear: "Thank you, but I can make my own decisions." What are your thoughts on the EU's regulatory approach to AI? How can we balance innovation and ethics without compromising our competitiveness? Share your insights in the comments below! #EUAIAct #Innovation #AIRegulation #Competitiveness #TechPolicy Christian Lindner Robert Habeck Sozialdemokratische Partei Deutschlands (SPD) CDU Deutschlands FDP Freie Demokraten Ursula von der Leyen Friedrich Merz European Commission Jörg Kukies

🚀 Exciting news from the world of technology! 🌟 Apple has recently announced its decision to withdraw from the self-driving car market to focus its resources more on the development of generative AI projects. This strategic move not only promises a fresh impetus for the landscape of innovation but also opens up new perspectives for the future. Here is an article that provides further insights: https://lnkd.in/eq-YpQ9F #Apple #Innovation #AI #GenerativeTechnologies #InnovationUpdate #AIAdvancements #FutureTech #TechIndustry #GenerativeAI #StrategicShift

Now, it should become clear what kind of headroom Nvidia, AMD & Supermicro - the whole AI hardware value chain - still are going to have for further growth. #SORA can „only“ generate 1minute of HQ video today - for a pixar style hollywood production massively more hardware, parameters and model size are required and there is no doubt, this is where it´s going, or maybe the next 30sec superbowl ad as an interim step - this is advancing faster than anyone can follow. Check it out for yourself Sora Intro - OpenAI Website https://openai.com/sora

Now this is interesting... OpenAI is expanding towards neuromorphic chips. In my opinion a bold first leap in the right direction to achieve AGI. OpenAI, under the leadership of CEO Sam Altman, has signed an intent agreement to invest $51 million in acquiring neuromorphic AI chips from Rain AI. These advanced chips, designed to mimic human brain functions, are expected to revolutionize AI training with a 100-fold increase in processing power and a 10,000-fold improvement in energy efficiency compared to current GPUs. Rain AI has already engaged in discussions with tech giants like Google, Oracle, Meta, Microsoft, and Amazon regarding chip sales. This strategic move aligns with Sam Altman's personal investment in Rain AI and highlights the convergence of his personal interests with his role as OpenAI's CEO. Additionally, the increased scrutiny of Middle Eastern asset funds in the U.S., particularly in AI-related transactions, is exemplified by the U.S. government's intervention in Prosperity7's sale of its stake in Rain AI. Despite the relative modesty of OpenAI's $51 million investment compared to Nvidia's soaring revenues, it underscores the growing demand for AI chips in the industry. #ai #neuromorphic #chips https://lnkd.in/defePxTb

Another month, another tech column in manager magazin. AI is still booming: GPT4o is pretty mind blowing and after publishing quarterly earnings, Nvidia is now worth $2.7tn (🤯) - more than Meta, Tesla, Netflix and IBM taken together and more than the entire DAX 40 (😪). There is little doubt that some AI companies are overvalued and will have to face a correction in valuations at some point. At the same time ✨ value creation is real ✨ especially at the hardware/infrastructure and foundation model layers, where winners are now emerging. As the application layer is still much more up for grabs, this is where we may see the sharpest valuation corrections in the future. Right now, users often hop back and forth between different tools, but if done right, e.g. through deep integration and vertical process understanding, applications also present a huge untapped opportunity for value capture. Whether Nvidia can defend their quasi-monopolistic market position (their market share in training is estimated to be over 80% 🤨) in the long run remains to be seen - competition never sleeps (hello Apple, Microsoft, Intel Corporation Groq, Recogni... many more) and especially the much more dynamic inference market will require significant innovation. Despite the current hype, AI will be the defining technological force of this century. Like many new technologies, we probably overestimate the amount of change we will experience in the short term yet underestimate the long-term transformative power. And so, in 10 years from now, Nvidia's current valuation may look quite cheap. 👉🏽 the full article is linked here: https://lnkd.in/ePAsuBpv See you next month!

It was a coup for Microsoft. Its boss Satya Nadella had made a personal effort to bring Mustafa Suleyman to the company. As one of the three co-founders of Google DeepMind, Suleyman is an entrepreneurial AI pioneer. Not only that: Suleyman, who founded Inflection AI two years ago, also brought one of his co-founders and a number of talented individuals with him to Microsoft. It is rare for a founder to move to a large corporation without their start-up being acquired as well. Microsoft is also one of Inflection AI's investors. However, the company is said to have shied away from the takeover for fear of delays caused by competition authorities. Another reason for Suleyman's move was lost in the coup. He had difficulties finding a plausible business model for Inflection AI's chatbot Pi. The main reason: the skyrocketing costs of training and using AI models. This is a huge problem in the industry. Dario Amodei, founder of Anthropic, expects that we could see the first AI models costing ten billion dollars to create as early as next year. According to the Financial Times, even Google is therefore considering charging for part of its AI search. However, this is unlikely as long as Microsoft is trying to make searching via Bing using AI more attractive and offering this free of charge. This WirtschaftsWoche story looks at how the industry is dealing with the cost problem and what is being done about it.

Sam #Altman, CEO of #OpenAI, should not only plan his own #AI chip, but entire fabs. Now he is looking for finishers or cooperation in South #Korea. Previously, it was already said that he was traveling in the Middle East, Arab donors are said to have been the target, among other things. AI, i.e. artificial intelligence, is on everyone's lips. Whether justified or not - it triggers a boom from which many want to benefit. After the drama about #OpenAI, #Microsoft and the person Sam Altman recently, they are now making speeches with new plans. Not only should your own AI chips be planned, but rather you would also want to produce the whole thing in your own fabs, at least in part. A meeting with industry giants in South Korea is to take place this week, previously it was said that the feelers had been extended towards #TSMC. https://lnkd.in/dTF-EFY7