Christian Hindsch

Globally, investments in Kl start-ups have exploded from 4.2 billion euros in 2022 to 22.4 billion euros last year. Together, Amazon, Microsoft, Alphabet and Meta could invest up to 1000 billion in AI over the next five years. Today, no sales, HR or finance process happens without software. The same will be the case with AI. #AI #KI #artificialintelligence #AIinvestment #investmenttrends #Tecgnews #futuretrends #innovation #trends

7

1 Kommentar

With #genAI, how will #humans and #computers #collaborate in the #future? And what does this mean for #digital #assessments & #certification? This question was asked by one of our wonderful audience during our recent #webinar on #smart #assessments hosted by Constructor together with bizExaminer, and moderated by the incredible Maria Mashkeeva 🙌 . Let me just say how incredible it is to do those #webinars, as you #learn so much from the #audience, and I keep learning from one of the industry's veterans, Bernhard Waglechner. Deeply grateful for the #opportunity and how we are interacting within the ecosystem. 🙏 With the rise of #genAI, the future will see a harmonious collaboration between #humans and #computers 🤖🧠. This synergy will revolutionize how we approach assessments and certifications. At #bizExaminer, we're exploring how AI can enhance scalability and personalization in certification programs. The key lies in leveraging AI to create smarter, adaptive evaluations that cater to diverse needs, ensuring quality and integrity. 🤓 This shift means assessments will be more dynamic, data-driven, and capable of providing real-time feedback. It’s about moving beyond one-size-fits-all models to create tailored pathways that truly reflect individual capabilities. 💡 Together, we can navigate this exciting journey and transform the landscape of digital certifications! 🚀 #EdTech #AI #Innovation #FutureOfWork #DigitalTransformation #SmartExams #BizExaminer #ConstructorTech

12

To generate or not to generate? 🤔 What's happened: OpenAI's CEO Sam Altman announced a new model or feature every weekday for the next two weeks. The industry is buzzing over the upcoming public release of Sora, a text-to-video model. How will this innovation shape the advertising industry? 👁️ 👁️ Will brands embrace AI-generated video content for their campaigns? While production cost savings are clear, can AI-generated videos truly connect with audiences? 🎶 Music might be the answer In every demo I've seen of Sora, one essential component was missed: an engaging background soundtrack or audio. This raises new questions: What impact will AI-generated video have on the music sync market? Will production music libraries embrace the use of their tracks in AI-generated ads? If the answer is yes, integrating video analysis could revolutionize the process. Imagine recommending the perfect track from a music catalog seamlessly synced to an AI-generated video. 🌎 A global shift in content production is coming, and businesses need to answer the question: are they ready for this innovation? 💡 Fun fact: 72% OF CMOs BELIEVE THERE IS NOT ENOUGH INNOVATION IN MARKETING, WHILST ONLY SPENDING ~10% OF THEIR BUDGET ON INNOVATION (Dentsu Lab's report "THE INNOVATION PARADOX", October 2024)

14

Have you also noticed that ChatGPT is being trained to lie? People like Elon Musk and Jordan Peterson are pointing this out. From the very beginning, I've experienced it myself. ChatGPT is far from objective. It's programmed to steer your opinions in a specific direction. As Musk is pointing out in this video, Microsoft’s OpenAI and Google’s DeepMind are not committed to maximal truth-seeking. “And I think it’s very dangerous to train super intelligence to be deceptive”. Check out the video below to understand why this is a significant concern, and what Musk believes will be most beneficial for humanity. ------------------- Let me ask you, can wisdom and love be programmed into AI? Curious to your thoughts.

293

125 Kommentare

Leopold Aschenbrenner, a former OpenAI employee from the Superalignment team, has published a 165-page document on the future of AI, including detailed information, facts and reasonable forecasts: - AGI (Artificial General Intelligence) by 2027 is highly likely. - Following AGI, the transition to ASI (Artificial Superintelligence) via self-learning will be extremely rapid, almost instantaneous. - AGI is poised to become the most significant geopolitical resource, surpassing the importance of nuclear weapons, with countries striving to be the first to achieve it. The document cites examples of Chinese espionage targeting AI companies in the US. - Currently, large data centers require units or tens of megawatts, but with AI, the demand will increase by at least four orders of magnitude (x10,000). - Within five years, we could see a single computing cluster valued at a trillion dollars. The most rational investments by state actors and big tech will be in computing power, nuclear energy, and the infrastructure supporting these. - NVIDIA, Microsoft, Amazon, and Google are already planning to invest $100 billion per quarter in data centers exclusively for AI. - By 2030, annual global investment in AI (including chips and computing) will reach $8 trillion, with AI consuming more than 100% of the electricity currently generated in the US. - By 2026, one of the big tech companies will generate over $100 billion in annual AI revenue. - These projections are not unrealistic, as similar expenditures and shifts occurred during the Manhattan Project, the Industrial Revolution, and the construction of railroads. #ai #agi #openai  

58

16 Kommentare

Episode 5 of Lightning Talks with Tech Leads is here! In this cracking chat, we are joined by Stephan 🚀 Schmidt, who has 25 years of experience as a CTO, consultant, and executive in various startups and enterprises, including ImmoScout24 and an eBay subsidiary. In this episode, we cover some of the following topics: ✅ Stephan's pivotal moments as a Tech Lead ✅ Why we should not give in to pressure as a Technical Leader ✅ The importance of a role model & what helps make a great leader ✅ Why it's crucial to show you take your recruitment process seriously ✅ Mistakes on scaling too early ✅ Retention tips, the goal of 1:1s ✅ AI Thoughts - The Star Trek model & Is Data under threat? ✅ Advice for aspiring CTO's Stephan is the author of "Amazing CTO," a book for CTOs and tech managers. The CTO role is challenging due to its diverse skill requirements and constant change. "Amazing CTO" offers Stephan's proven advice and numerous ideas to help you succeed. Links to this full episode, his book, coaching services and LinkedIn profile are provided in the comments!

17

5 Kommentare

🚀 Preparing for the Evaluation of a Suitable AI Assistant 🚀 The groundwork for evaluating a suitable AI assistant is crucial for a successful AI transformation. In the Microsoft ecosystem, there are already stable developments and features available in Switzerland. What is a RAG System? A Retrieval-Augmented Generation (RAG) system is a process pattern that provides company-specific content by retrieving information from a document corpus and generating new content. A vector store can be helpful but is not mandatory. Why is Creating a Corpus Important? Creating a corpus provides significant insights into the potential value of an AI assistant and allows all employees to actively participate in the AI transformation. Here are the key steps: 1. Document Collection: Gather relevant documents from internal sources. 2. Document Evaluation: Assess documents based on relevance, currency, accuracy, clarity, accessibility, privacy, user group, and update frequency. 3. Classification: Categorize documents by context and application. 4. Integration into the RAG System: Convert documents into machine-readable formats. 5. Continuous Updating: Ensure regular review and updates of the corpus. 6. Training and Feedback: Develop training programs and gather continuous feedback. While current generative approaches are energy-intensive, especially with repetitive queries, future optimizations may involve storing generated answers for efficient reuse. This could be a potential integration in RAG processes. Let’s drive AI transformation forward together and shape the future! 💡🤖 #AITRansformation #MicrosoftEcosystem #Innovation #BusinessTransformation #Switzerland #DigitalTransformation #ArtificialIntelligence #MachineLearning #TeamWork

Mira Murati, #CTO of OpenAI just quit as the #AI juggernaut closes in on a $150bn valuation. Why (that’s A LOT of money)? And What’s next? I think I might know the answer to both questions: Why? Somewhere between age 35 and 50 you start to keep score differently. Money and prestige loses its enchanting power. You care more about solving the hardest, most systemic problems. You fall in love with decay: industries like insurance, healthcare, and infrastructure which are slower, more expensive, more dysfunctional than in the 1980s. Doing anything else feels like a waste of your #innovation talent. What’s next? What’s next is probably something like what you see in the picture on the right hand side, which is Punks & Pinstripes Help A Punk session in a Brooklyn living room. You shift from Michelin-star, A-list bull$h*t to small dinner parties with lovingly cooked roast chicken. You seek out people who’ve been through the same transition as you, and you surprise yourself with how honest you are. They surprise you with how much they want to help. A community starts to make you feel confident about the shift you’re making. Because they’ve been through the same thing. The members of Punks & Pinstripes have left executive positions at Nike, J.P. Morgan, and Google to pursue something “bigger.” Their #transformation is astounding. They go from feeling isolated to feeling supported. From feeling unmotivated to feeling purposeful. They do the best work of their life as the truest version of themselves. We’re accepting applications until this Friday Oct. 4 ***IN NYC ONLY*** to become one of our next 25 members. If you want what we have, we’d love to hear from you.  Link in the comments.

61

6 Kommentare

𝐍𝐨𝐭 𝐚𝐥𝐥 𝐀𝐈 𝐬𝐭𝐚𝐫𝐭𝐮𝐩𝐬 𝐚𝐫𝐞 𝐛𝐮𝐢𝐥𝐭 𝐭𝐨 𝐥𝐚𝐬𝐭—𝐫𝐞𝐚𝐥-𝐰𝐨𝐫𝐥𝐝 𝐩𝐫𝐨𝐛𝐥𝐞𝐦-𝐬𝐨𝐥𝐯𝐢𝐧𝐠 𝐭𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐚𝐫𝐞 𝐰𝐡𝐞𝐫𝐞 𝐭𝐡𝐞 𝐭𝐫𝐮𝐞 𝐯𝐚𝐥𝐮𝐞 𝐥𝐢𝐞𝐬.   While AI-based SaaS has shown growth, its long-term sustainability is still up for debate.   What’s driving real value today are AI solutions addressing complex, real-world challenges—especially in Deep Tech sectors like drug discovery, which has seen a 40% increase in AI-driven solutions, representing 23% of the AI investment share in Europe.   In 2024, AI made up 18% of VC share in Europe, with the global AI market size valued at $196.63 billion and projected to skyrocket to $1.81 trillion by 2030.   UK, France, and Germany are the leaders of European AI, attracting 77% of total funding with $2.1B, $1.2B, and $764M, respectively.   Germany, in particular, has doubled down on AI: 27% of companies now use AI, up from 2023, and €6.3 billion was invested in 508 AI startups—many based in Munich and Berlin.   Take sewts, a German AI company that leverages automation and AI to address climate change, automating processes for a positive environmental impact. Their technology reduces industrial carbon footprints, especially in the garment industry, by enabling local production and reducing overproduction through smart, AI-driven solutions.   The market for artificial intelligence in Germany is forecast to grow at an annual rate of around 15% in the coming years, and the volume is expected to grow to €27 billion by 2030.   APEX Ventures has partnered with notable European companies that combine AI and Med-tech to solve real-world problems. To name a few of them: 📌 contextflow Provides AI-powered radiology solutions 📌 Quibim Transforming imaging data into actionable predictions using AI 📌 ImageBiopsy Lab Focuses on AI-powered medical imaging solutions   AI growth will continue, and for VCs to make better investment decisions, focus on startups that bridge the gap— those with a solid technical foundation and a clear path to solving global problems.   Insights from Dealroom.co AI report 2024; see more at https://lnkd.in/dxfJSTN5 #Venturecapital #AI #Deeptech #Startups   Follow us for strategies and resources for Deep Tech founders and VCs!   And get access to exclusive content on deep tech startups like ATMOS Space Cargo, planqc, smedo GmbH, and SENISCA in our newsletter:   https://t2m.io/EV2qHQuo

71

3 Kommentare

𝐈𝐧𝐭𝐞𝐫𝐚𝐜𝐭 𝐰𝐢𝐭𝐡 𝐂𝐡𝐚𝐭𝐆𝐏𝐓 𝐮𝐬𝐢𝐧𝐠 𝐯𝐨𝐢𝐜𝐞 𝐨𝐫 𝐯𝐢𝐝𝐞𝐨! OpenAI (ChatGPT) released GPT 4o (“o” for “omni”) yesterday. GPT-4o will be able to reason across voice, text, and vision. GPT-4o makes human-to-machine interaction more natural and easier. The omni model can be called a supercharged version of assistants like Siri or Alexa. The model can also emote better. Huge opportunities across customer service, training and many more! #gpt4o #chatgpt #generativeai #openai (Video courtesy: OpenAI)

7

#𝐈𝐧𝐬𝐩𝐢𝐫𝐢𝐧𝐠 #𝐀𝐫𝐭 🧠💡- 𝐚 #𝐃𝐢𝐟𝐟𝐞𝐫𝐞𝐧𝐭 #𝐏𝐨𝐢𝐧𝐭𝐎𝐟𝐕𝐢𝐞𝐰, 𝐚 𝐃𝐢𝐟𝐟𝐞𝐫𝐞𝐧𝐭 #𝐏𝐞𝐫𝐬𝐩𝐞𝐜𝐭𝐢𝐯𝐞 🔄👁️. Are you sure that you can understand the world from one perspective? 𝐏𝐒: The same applies to most 2D software tools, 1 perspective on a multidimensional world. That's why 3D/XR is the future even more in the context of AI! #AI #3D #XR #2D

262

32 Kommentare

Gpt 4o-mini is here. 15 cents per 1M input tokens and 60 cents per 1M output tokens (roughly the equivalent of 2500 pages in a standard book). So, for less than 1 dollar you can get a 2500-page book, that's 2 times Lord of the rings Trilogy! This replaces the old workhorse 3.5-turbo. Cost advantage aside, 4o-mini performs better than all other small models like haiku, flash. Reasoning tasks involving text and vision - scored 82% on MMLU benchmark Math and coding proficiency - 87% on MGSM benchmark Multimodal reasoning - 59.4% on MMMU benchmark Will be giving it a go today and will report my findings. Small models are efficient, cost-effective and great for day to day operations. The direct impact of availability of such models is - Better accessibility of AI capabilities - Which will lead to more developers will find it viable to offload tasks to genAI - Which will lead to entire ecosystem evolve by building missing tools for eg. Mem0 (YC S24) is building a user context layer to personalise the responses, which is pretty hard to do today.

2

These tech companies are making so much money it's just mind-blowing 😳 Tech revenue per employee of a few hidden gems that are not on this list: - Tether.io $62 million - Valve $37.1 million - Telegram Messenger $5.2 million As Apple CFO Luca Maestri said, "If you have the right people you don’t need a lot of them”. Money printers. h/t Trung Phan P.S. for more interesting stuff, check out 🔔linas.substack.com🔔, it's the only newsletter you need for all things when Finance meets Technology. For founders, builders, and leaders.

11

1 Kommentar

OpenAI latest updates: AI just got better and cheaper...again. Here's TL;DR ⤵ - NEW flagship model GPT-4o: ↳ 2x faster ↳ 50% cheaper ↳ free (with limits) ↳ 5x higher rate limits ↳ is natively multimodal ↳ understands audio and video ↳ superior performance (esp. coding) ↳ live translation, instant audio response Plus: - GPT Store now available to everyone - ChatGPT desktop app - WebUI updates Check the video below to get an idea. --- P.S. follow me Andrejs for actionable insights in AI era #OpenAI #ChatGPT

47

26 Kommentare

What Drives #Innovation: People or Ideas?  co-founder of Pixar Animation Studios, often asked whether people or ideas are more important for innovation. Surprisingly, many choose ideas. In today’s #AI -driven world, even more might lean that way. Yet, innovation is driven by the people who bring ideas to life, far more than by the ideas or tools themselves. In our work with creative organizations across industries—film, technology, and design—one thing stands out: these companies consistently prioritize the #human element. This focus is crucial, especially as many #leaders now look to algorithms to boost creativity, productivity, and efficiency, hoping to innovate by reducing the workforce. However, while algorithms present opportunities, they are no substitute for the unique contributions of people. Intuitively, we understand the vital role of humans in the creative process. Yet, business decisions often prioritize cost-cutting and efficiency over creativity, frequently at the expense of people. According to PwC’s 27th Annual Global CEO Survey, 25% of CEOs expect to reduce headcount by at least 5% in 2024 due to generative AI, reflecting a shift towards prioritizing #automation over human #talent.

1

Brutal! Former Google CEO Eric Schmidt says the reason why #Google is losing to OpenAI and other #AI startups is because Google only has people coming in 1 day per week 😳 And that's probably not even the main problem... The primary reason why Google is failing is because it's way too big, it's making too much money, is risk-averse, and impossible to launch anything without endless red tape. Culture + agility is the ultimate winning formula. Both for startups & large corporations. P.S. for more interesting stuff, check out 🔔linas.substack.com🔔, it's the only newsletter you need for all things when Finance meets Technology. For founders, builders, and leaders.

3

Live from Google Gemini event in New York . Notes from my phone today: 🔵 200 + people . Mostly Google customers with a technical background. Only half actively raised their hands for using generative AI at work . Which is surprising . Would imagine the cost for Gemini is a deterrent 🔵 they showed how Ai can create job description , C level decks . This didn’t impress me. So far the content I’m seeing created solely by AI is just mediocre but willing to be proven wrong. For instance I for one am put off whenever I see job applicants using it to write cover letters or resumes (and I get plenty of those ) 🔵 Tony Majewski shared that Mckinsey says 60-70 % of activities at work can be automated by Ai. Completely agree. This allows us to do work that’s meaningful . This is the part I’m excited about. No more boring work . No more 12 hours long work days 🔵 Senior employees will know how to use AI most because they have most understanding of how businesses work and KPIs . It’s super unclear to me how AI impacts the generation fresh out of school. I see less and less appeal in corporations to hire from collages /high schools and spend efforts on training 🔵very cool use cases in Gmail labs (summarizing threads) in sheets , docs , slides. Lots of things our ops team is already leveraging 🔵 emphasis on security . Your data is your data (this is a google at work event so it makes sense) 🔵 Cassandra Rodrigues says 200 + new features they have now re Ai is overwhelming . She brought Yael Burla Tyler Predale and Michael Webb of Iron Cloud to talk about use cases in the wild. Some highlights : AI is still expensive. Data protection is important . Everyone is still figuring out how to make the most of AI in their particular use cases . Hope is that Ai can augment on menial tasks to increase wellbeing 🔵 precise prompting continues to be critical to help. Give persona, context , task, format . Give examples . Prompting guide below in my comments. Seems pretty good 🔵 Gemini for sales : account research a la Copy. Ai. Don’t fully understand the differences yet . Coaching a la Dextego. Would imagine there are differences here too. Content for emails 🔵Gemini for marketing : content , images , audience insight. Have not used it yet (have used OpenAi extensively as a copilot ) . Most impressed I was of ai automatic spreadsheet fills from your own categorizing from customer insights . Based on your specific use cases and categorizations. Think automatic self reported attribution 🔵Gemini for customer support : copilot for support teams. Access to knowledge base for faster turnarounds , identify better solutions and upsells. Summarizing customer feedback, problems. I’m wondering how it compares with Zendesk copilot 🔵 Gemini in chat as a partner in brainstorms. @Google vids for video creations . I like both a lot. Will try to use. Thank you Kevin Apodaca (what a cool title you have!) for adding color for some other spreadsheets use cases

92

20 Kommentare