himala AI

?? ?Official Reveal – Meet himala??? Today, we say goodbye to #Oxolo and hello to himala. This is more than a name change—it’s a step forward into what we’ve been building together: a digital partner designed to bring balance, clarity, and focus to your life. We created himala to help you cut through the chaos and focus on what truly matters. From streamlining your daily tasks to helping you achieve long-term goals, himala is here to make life simpler, clearer, and more intentional. This is just the beginning, and we’re so excited to share this journey with you. Welcome to himala.??? #LaunchDay #hellohimala?#GoodbyeOxolo #NewEra

?? The Myth of Multitasking ?? We’ve all been there—juggling a dozen tabs, flipping between emails, calls, and to-do lists, thinking we’re being productive. But here’s the truth: multitasking often just adds noise to the day, scattering focus instead of sharpening it. We’ve been rethinking how to solve this challenge and create something that doesn’t just help you manage your tasks but helps you manage your focus. Stay tuned, something new is on the horizon… ?? #Focus #ComingSoon #Productivity

?? Imagine Life, Seamlessly Managed ?? What if you had a human-like AI companion, assistant or co-pilot that truly understood what matters most to you? Picture fewer interruptions, more focus, and a sense of balance that keeps you grounded, no matter how busy life gets. With something exciting on the horizon, we’re thrilled to introduce a solution designed to simplify your day—your way. ??Stay tuned for a smarter, more intentional way to manage work and life. ? #Teaser #FutureOfWork #AI

?? Big Changes Are on the Horizon??? After weeks of intense brainstorming, planning, and more coffee, matcha, energy drinks… than we care to admit, our team is excited to share that a transformation is underway. ?But here’s the truth — it hasn’t all been easy nor perfect. This pivot meant making tough decisions, and we even had to say goodbye to some incredible people along the way. It was hard, and it made us pause and reflect. Yet through it all, we turned those challenges into a positive force, driving us to reimagine our path with even more purpose. The past three months have been a journey of growth, grit, and gratitude. We’ve shaped an idea we believe will elevate how we all interact with technology and productivity. ?? ?Stay tuned... Something new is coming, and we can’t wait to introduce you to what’s next. #ComingSoon #NewBeginnings #PivotingForward

???? ???? ???? ?????????????????? ?????????? ???????????????? ?????? ??.??. ???????????????????????? ????????????????? ???????♀? History is happening right in front of our eyes. Taylor Swift is endorsing Harris and had Trump not posted the AI generated Image of Taylor Swift falsely endorsing him, she probably would not have publicly supported Harris. ?? "???????????????? ?? ?????? ???????? ?????????? ???????? ???? ???? '????' ?????????????? ?????????????????? ???????????? ??????????'?? ???????????????????????? ?????? ?????? ???????????? ???? ?????? ????????. ???? ???????????? ???????????????? ???? ???? ?????????? ???????????? ????, ?????? ?????? ?????????????? ???? ?????????????????? ????????????????????????????. ???? ?????????????? ???? ???? ?????? ???????????????????? ???????? ?? ???????? ???? ???? ???????? ?????????????????????? ?????????? ???? ???????????? ?????????? ?????? ???????? ???????????????? ???? ?? ??????????. ?????? ???????????????? ?????? ???? ???????????? ???????????????????????????? ???? ???????? ?????? ??????????. ?? ???????? ???? ?????????????? ???? ???????? ?????? ???????????? ???????????? ?????? ?????? ???????? ???? ?????? 2024 ???????????????????????? ????????????????. ??'?? ???????????? ?????? @???????????????????????? ?????????????? ?????? ???????????? ?????? ?????? ???????????? ?????? ???????????? ?? ?????????????? ???????? ?? ?????????????? ???? ???????????????? ????????." There is a certain asymmetry in their relationship- Trump clearly likes Swift, she's everything he admires: rich, beautiful, successful and good at business. Yet, he's everything she resents ("those aren't your dad's republicans") but she still would probably not have jeopardized losing fans over politics- it seems like it takes a lot to irritate her- but Trump's post on X got the job done and now she might decide the election. Her predominantly female fan base called Swifties are regarded by journalists as one of the largest, most devoted, and influential fan bases and are very likely to listen to their idol. And female voters will play a pivotal role in determining the outcome of the election. Women comprise the largest group of registered voters, and they tend to turn out at?high levels to vote, as 68.4% of them did in the 2020 election. ?The fact that an AI generated image triggered this cascade of decisions worries me though. What do you think? ??????????????: ???????????? ?????????? ?????????????????? ?????? ?? #uselection #genai #deepfake #catlady Dr. Jan Philipp Burgard

?????? ???????? ???????????????????????? ???????????????????????????? ?????????? ?????????????????? ???????????????????? ????????? ??Ilya Sutskever just raised a billion USD to fund his start-up SSI focused on safety. Clearly people are worried about AI being in the wrong hands and misinformation, manipulation, and distortion of truth. This is especially concerning when deployed in political contexts, where the stakes are incredibly high. ???? As we approach the next U.S. elections, the role of AI-generated images and deepfake technology becomes even more relevant and with great innovation comes great responsibility. But how much responsibility should be carried by the enterprises? How much should the regulator cover? Where should it be regulated? Do the individuals writing the code carry responsibility? How high should the guardrails be? If all sides have equal access to the same tools is it still a weapon? I don't have answers yet but would love to hear your thoughts. ???? ??????????????: ?????? ?????? ???????????????? via Jan Andresen #AI #Deepfakes #Geopolitics #EthicsInAI #US2024 #ElectionIntegrity

???????? ???? ?????? ???? ?????????? ?? At Oxolo we have been building Gen AI long before it was fashionable. We have built software for people to easily generate videos featuring humans. For e-commerce, for training and HR purposes. We have done it successfully, we’ve raised Series A from top tier VCs, we have over 600k users, have built a high caliber team and grew daily. ? However, we decided to discontinue the product oxolo as it is. Why? ? ?? ???????????????????? ???????? ?? ?????? ?????????? ???? ?????????????????????? ?????????????? ???????????????????? The AI Act came into place on August 1. Our avatars very likely fall under the “deepfake category” as they are a “AI manipulated video resembling an existing person”. Providers of AI systems generating synthetic content have to ensure that the outputs of the AI system are marked in a machine-readable format and detectable as artificially generated or manipulated. So imagine marketing videos with labels saying “AI GENERATED” on it. That’s not going to work. Further imagine the legal work behind trying to comply with an obscure piece of regulation. ?? ???????? ???????????????? ???????? ???? ???? ?????????? ???? ?????? ???????? There is a lot of insecurity about how to implement AI and also regulatory uncertainty. It will take a while until this will change. A recent study showed that 77% of employees report that AI has increased their workloads and hampered productivity (link in comments)- this of course will change and is only a matter of time. However, as a start-up we don’t want to wait it out. ?? ???? ?????? ???????????? ?????????????? ???????????? ???????? ???? ??????????????????- ?????? ??????-?????????????? Even though we believe there is a large market for lip-synched generated content - take one look at the video below (Source: Flux/Leo Kadieff). It’s only a matter of months until hyper realistic humans with the ability to speak will simply be generated. Without IP restrictions and AI Act compliance issues. ?? ?????? ??????? We didn’t take this decision lightly but timely. We will now build a product using our existing team and tech - most likely B2C. It’s been hard on us - letting go of people and a product and all the hard work that went into it is not easy. I want to thank everyone for holding up the fort - you guys rock. Thanks to our loyal customers - you were the best and thanks to our investors for your trust. ?? Onwards and upwards! ?? Heiko Hubertz ?????? Victoria Busse Gülsah Wilke #ai #genai #product #growth

?????? ???????????????? ?????? ??????????-???????????????? ?????? ?????? ???????? ???????????? ???????????? ???? ?? Meta's Lama 3.1 405B launched and every new model release opens the same discussion: ?????? ???????? ?????? ???????? ???????? ?????? ???? ???????? ???????? ?????? ??????????? ??? ?????? ?????????????? If we ignorantly assume that the quality of all models will more or less be the same (some feel a little anodyne but that will improve)- from a consumer perspective open source is the better option: ?? ?????????????? ???????????????????? ?????????? Enterprises don’t want to depend on models they cannot run and control themselves. Many organizations handle proprietary data they need to secure and can’t send over to closed models through cloud APIs. Other organizations simply don’t trust the closed model providers with their data. ???????????????????? ???? ??????????????. ?? ???????????????????? ?????????????? ???????????? Enterprises also like to customize. They can continue training open source with their data and then distill them down to a model of their optimal size – without anyone else seeing your data. ?? ???????? ?????????????????? Developers can run inference on open source Models on their infrastructure at roughly half the cost of using closed models, for both user-facing and offline inference tasks. ?????????????????????? From a societal perspective, AI has more potential than any other technology to improve our lives and an open AI ecosystem would allow access to a critical infrastructure and can be amended as needed. It's unlikely AI companies with only closed source will survive. ?? ???????????????????????????? However, is releasing the model without releasing the training data set (and code base, data methodology, and the weighting/normalization process etc.) ever open source? The move to release an open source model by a company ?????????? ???????? ???????????????? ?????????? ???? ?????? ?????????????? ???? ?????????????-while others are raising billions- is a statement which will be very hard to replicate. Meta has access to the biggest data pools which keep growing - 30% of the global population is using a meta product daily. it will cost Meta 10x to train the next model- Lama 4. It looks like Meta is loading up on capex to release more best-in-class open source models to flatten its competition. With similar data and distribution access, Google will probably be the only one to survive in the long run. So hopefully some competition will remain to prevent model-monopoly. No wonder OpenAI keeps panicky "pre-releasing" a whole variety of products (SORA, SearchGPT) to find streams of revenue they can compete with in the future. ???????? ???? ?????? ??????????? Felix Schlenther Tristan Post Rik van der Woerdt Jan Andresen Oxolo #genai #tech #datascience PS: The prompt for the video was Mark Zuckerberg, Sam Altman and Sundar Pitchai competing in the swimming Olympics on Kling. Not 100% satisfactory.

???????????????? ?????? ???? ?????? AI has the potential to completely disrupt how kids will learn and according to UNICEF kids are coming in contact with AI through their devices daily. Screen time and AI are deeply intertwined and woven into the fabric of daily life. ?? AI is in every application. Most interactive toys for younger kids depend on AI and depending on their age they might already have a digital footprint: search history, downloads, viewing, listening, AI-powered smart speakers, and the list goes on. But there isn't any comprehensive regulatory framework or even consideration of how AI will affect the well-being of our children. ???? ???? ???????? ???????? ???? ???????????? ?????????? ?????? ???????????????? ???????????????? ?????? ???? ???????? ????????????. ???? Social media gets smarter and more sophisticated with AI. Systems are being optimized to max out the dopamine release in their growing brains by individualizing content through self-learning systems that understand the mechanics and habits of the viewer. Children's brains are particularly vulnerable as their brains release more Dopamine than grown-ups- which means that even little spikes in the neurotransmitter from a piece of candy or a little “dingding” at Candy Crush feels really really good to them. ??In a recent study by the University of Edinburgh, children between 6 and 11 were surveyed about smart speakers, essentially chatbots with a voice. 75% thought that the systems might think like humans, most believed the systems were smarter than themselves, and 30% believed that AI could think for itself. 70% said it is wrong to be rude to the devices. Most three to six-year-olds believe that chatbots have thoughts, feelings, and social abilities. ???? Kids are easy victims. They share personal information, they trust AI more than they trust humans. We have to educate them early on the perils, we have to educate ourselves (only 1 in 4 parents knows if their child uses gen AI), we have to adapt to “anti-dopamine parenting” and we have to vote politicians in office who intend to set up meaningful regulations. What do you think can we do to make sure AI is not harmful? #ai #genai #coppa #socialmedia? Oxolo Fredrik Harkort Jan Marius Marquardt

?????? ?????????? ???? ???????????????? ?????? ???????????? ?????? ???? ???????????????????? ???? The scene in the video portraits a large group of China's society and is a reflection of a sentiment many have: more than one third of the Chinese labor are rural migrant workers, ranging at 297.5 million in 2023 (Statista) and a third of their kids "are left behind" (UNICEF). The video also shows how China's Gen AI scene is taking off and If the video serves as a proxy it shows how much further their tech is compared to the rest of the world. The relative maturity of the Gen AI market also might serve as a precursor for the U.S. and Europe- or will at least demonstrate what works and what won't for us in the future. Not only is China leading in regulatory aspects. They are leading in R&D, in cloud, training data, models and products and most importantly: China is leading in worldwide Gen AI ???????????????? in the enterprise sector which is propelling the AI revolution. ?? ??&????????: Last week, a report by the United Nations' World Intellectual Property Organization showed China was leading the GenAI patent race, filing more than 38,000 between 2014 and 2023 against 6,276 filed by the U.S. in the same period. ???????????? ?????? ???????????????? ??: Unlike Germany China manages to convert science into business. While the leading generative AI companies like OpenAI, face restrictions in China, the country has developed a robust domestic industry, with their own models and products from tech giants such as ByteDance. Just take a look at the video- not even SORA from OpenAI can compete. ?????????? ??: The cloud market is already consolidating and a price war is now heating up. As a result Chinese tech giants Alibaba and Baidu are slashing prices on LLMs. Alibaba's cloud unit announced price cuts of up to 97% on a range of its Tongyi Qwen LLMs. Its Qwen-Long model, for instance, will cost only 0.0005￥ per 1,000 tokens - or units of data processed by the LLM - after the price cut, down from 0.02￥ per 1,000 tokens. Bytedance announced last month that the main model of its Doubao LLMs would be priced 99.3% lower than the industry average for business users. ???????????????? ??: According to a survey of 1,600 decision-makers in industries worldwide by U.S. AI and analytics software company SAS and Coleman Parkes Research, 83% of Chinese respondents said they used generative AI. The global average in this study is 54%. The industries surveyed here included banking, insurance, healthcare, telecommunications, manufacturing, retail and energy. ???????????????? ?: Conventional wisdom suggests that technological breakthroughs proliferate in an environment of free markets, free speech, and democracy. So if you want to innovate, you need to operate like the U.S. or Europe- but the this does not seem to work anymore- we are losing. What do you think do we need to do to "retain" competitiveness? #genai #ai #china #unitedstates Oxolo