Eissa Hatem’s Post

Reflections #buildinpublic When to start integrating new AI tools into a fresh project? Back in 2021, I launched my first AI startup. I connected with a brilliant engineer who specializes in supercomputers and has been in the AI field for years. He provided me with essential reading on AI technologies (they’re not exactly new). I read through it all, and he clarified what I didn’t understand. Then, we started our work... manually. No matter how advanced we get, the key to any business remains the same: the customer and their needs. So, we developed a product and reached our first real sales, but we always kept in mind how automated tools could eventually take over. Why bother with AI if everything can be done manually? AI is crucial for drastically reducing product costs. Our startup’s goal was to make a luxury, expensive handcrafted product available to the masses. Since then, the AI and AI startup boom has taken off. But when I think about how to apply it, I still prefer to start with customers and manual work. Maybe I’m old-fashioned. :) How about you? Do you prefer an AI-first approach?

"The winner of the race I’m going to describe will earn billions. Maybe tens of billions. They’ll bend the arc of the universe. They’ll materially increase GDP. This is the race toward the AI agent. Agents are the next step in the AI race and the focus of every major tech company, research lab, and leading AI startup." - Every Newsletter - Evan Armstrong @ItsUrBoyEvan What Are AI Agents—And Who Profits From Them? The newest wave of AI research is changing everything https://lnkd.in/gMRZu33y

I have the feeling the way of doing startup and looking for PMF has completely changed, right before my eyes, in the past 2 years. 🤯 We tested generative AI at Voltaage, and it totally disrupted our processes. In a nutshell, at the beginning our job sounded like this: 👉 We've found a problem 👉 We evaluate the market to solve that problem 👉 We sell an offer 👉 We create a crappy solution If the same thing repeats itself 50 times in the last two steps, we change "crappy" to "good." ✅ If you're lucky like me, Edoardo Carlevaris and Domenico Scalera, you were already working on that problem and industry beforehand, and it's slightly easier (no big deal easier but still... 😅) Now, imagine something gets in the mix while the cycle wheel is spinning: ⚡ The channel and the way people ask questions changes in a few months ⚡ Your product thinks about improvements before you even see the questions ⚡ A "crappy" solution for a new facet of the problem is proposed in a single feature before you can create a new offer 😳 For us, in the end, it was a change for the better 🙌, but I wonder if that's always the case and if there’s any new literature on this. #ProductMarketFit #AI #AIstartup #data

Is consumer facing AI still a promise? Let me come clean, I truly believed that the Humane pin and Rabbit R1 would revolutionize the product design landscape. The user would be freed from the shackles of interacting with endless interfaces, now only having to give a simple command. I should have know better. After watching numerous disastrous videos on YouTube, I can only conclude one thing: We aren't there yet, and it doesn't seem like we'll be there anytime soon. Getting LLM right is just incredibly challenging. So what comes next? Do we have to wait until the next startup succeeds after several have failed? For now, it looks like big tech is going to pick up the tab. They have the money, energy, and platform dominance to push their policies forward. However, to make AI successful as a consumer-facing product, big tech needs to overcome two major challenges within: a system that delivers truly helpful results by understanding the full context, and communicates as naturally as talking to a person. As of now, we have seen glimpses of what may come. For the time being, I entertain myself with less potent forms of artificial intelligence, such as Siri, which, after 13 years, promises to finally help with your tasks instead of just providing references to websites. Let's see. #productdesign #AI

🚀 Building the Future of Material Selection with AI: A Journey of Innovation and Challenges 🚀 Hello LinkedIn community! I’m thrilled to share some updates and insights from the development of my AI-driven tool, designed to transform material selection and manufacturing processes. As many of you in the AI/ML space know, launching a product in this field is both exciting and challenging. Right now, I’m navigating a crucial aspect of our project: balancing our dataset to ensure our model delivers accurate and reliable material recommendations across all use cases. The Challenge: In our current phase, we’re working to ensure that our model can accurately recommend materials across a broad spectrum of applications. However, some of our material classes are underrepresented in the dataset, leading to potential bias. This is a common issue in machine learning but one that requires careful handling to ensure consistency and generalisability. Why This Matters: For our AI tool to become the industry’s go-to solution for material selection, it needs to perform exceptionally well across all scenarios, not just the most common ones. This means finding a way to either enhance our data intelligently or collect more diverse data points to train our model effectively. The Questions I’m Pondering: • Should we continue using automated techniques to balance our data, or is it time to manually enhance our dataset? • How can we ensure that our MVP not only works but consistently delivers value across diverse applications? • What strategies have you found effective in overcoming similar challenges? Your Insights: I’m reaching out to this incredible community for your thoughts and suggestions. Have you faced similar challenges in your AI projects? How did you overcome them? Your input could be invaluable as we push forward to the next milestone. Follow me for more updates as we continue to innovate and bring this tool to market. Let’s build something incredible together! Looking forward to your comments and ideas! 👇 #AI #MachineLearning #TechInnovation #StartupLife #DataScience #MaterialScience #ProductDevelopment #Innovation #Entrepreneurship #ArtificialIntelligence #FutureOfWork

So, you want to make AI your sidekick in MVP development? Smart move! But remember, even Iron Man had to learn to fly his suit without smashing into walls. AI is a tool, not a magic wand. Get this: A startup used AI to predict user behavior and bam! They cut down feature development time by half. Want to try? Spend 5 minutes googling an AI product relevant to your project. It’s like choosing the right superhero gadget — you gotta know what each button does! After choosing a good enough product, take some time to learn how to get the best out of it. You’d be surprised how far you can go with a little bit of better understanding. Don’t put all your eggs in one AI basket. Research, Learn, Build. That’s the way to go.

I am excited to finally launch the Makir.ai Marketplace on Product Hunt! We’ve been working hard to build a platform where AI startups can get discovered, and creators can find tools that fit their needs. Looking forward to the feedback feedback as we continue to improve and grow the community! #ProductHunt #Makir #AI #Startup #LaunchDay

🚀 We’ve just launched on Product Hunt! 🚀 We are thrilled to introduce Makir.ai, a project born from our passion for discovering and experimenting with AI tools. The AI space is booming, but with rapid growth comes the challenge of finding the right tools amidst the chaos. That’s where Makir.ai steps in: 🔍 Discover the Best AI Tools – Our marketplace makes it easy to find, compare, and implement AI services that fit your needs. 🌟 Showcase Your Service – Customizable listings help developers reach the right audience, fast. 📌 Embeddable Cards – Seamlessly integrate your service listings into your own website. 🔥 Featured Listings – Gather feedback, get noticed, and earn a spot on our home page. 💡 Canvas Workflow Creator – Automate workflows with natural language, and let Canvas design the perfect solution. We’re live and in Beta, and during this phase, Makir.ai is completely free to use! We’re excited for your support and feedback as we continue to build the go-to platform for AI innovation. Let’s make this launch a success together! Check us out on Product Hunt: https://lnkd.in/gY4jBxKS 💬 Drop by, leave your thoughts, and let’s shape the future of AI. #ProductHunt #Makir #AI #Startup #LaunchDay

I spent ZERO minutes creating these stunning images! ⏱ Swipe to check out the results! 👉 If you're spending time perfecting your prompts, then you're defeating the purpose of AI altogether. Why? Because AI should understand you, not frustrate you. I was able to create these amazing images with zero prompt engineering—just shared my vague thoughts, and it was able to understand it all. Want to know how I did it? Just comment "Fibr" and I'll DM you the hacks. #AI #Marketing #Startup

More than half of the AI focus companies that started in the last 24 months will face this obvious demand from their customers and their clients and those without business models or revenue streams that are really apparent... Will fail... And if they have good technology it will be sucked into the larger players.. And if they were making up things well they'll just make us all look bad. Some of the ones that will survive will have chosen to solve hard problems and given semi-realistic timelines to their investors.. I'm talking about people that have raised two or three hundred million and knew that he had 5 or 10 years in front of themselves to build things out and do experimentation in a professional manner that happens in many software startups. As some of the companies start failing, there will be many foolish skeptics that will show up in the newspapers who really aren't part of the industry anymore that will claim that this is the end of the technology or that yeah it doesn't work so we should stop things... That literally is ridiculous. In this stage of software evolution where there're already many proven ideas that have good enough ROI to start using generative artificial intelligence, machine vision and machine learning eg to optimize expenses and revenue in companies. AI doesn't do everything, but it does do somethings well at scale. good article... I recommend to grok it. #ai #iaadvice https://lnkd.in/g_H3etvF