Volvo Group’s Post

The #Cybertruck owner manual has more than 121,970 words, which would take 15+ hours to read at an average pace. See how #Qvio uses #controlledAI to deliver Tesla's answers in real-time. https://lnkd.in/g2ipfnsg

Since Tesla released their FSD12 model and announced it was using End To End Learning, I got tons of questions about E2E. And in particular, about the « Black Box » problem. Why did the car turn left? Stopped suddenly? What’s driving the decision, if all is a Neural Network. First, it’s not one neural network, it’s not 2017 anymore, and the best E2E models are now several models plugged. But also: In this morning’s daily email, I shared 5 techniques to understand E2E models: 1️⃣ Attention Map Visualization 2️⃣ Auxiliary Tasks Visualization (objects, depth, segmentation, …) 3️⃣ Cost Map Visualization (occupancy, drivable area, … directly impacting the driving decision) 4️⃣ LLMs! Car-GPTs can explain why a model is doing what it’s doing 5️⃣ Uncertainty Visualization, like Alex Kendall ‘s awesome work on Epistemic Uncertainty with Wayve Wanna see a bit of it? Here’s an end to end model, and the visualization: #deeplearning #selfdrivingcars #robotics #computervision #tesla #ai

Jérémy Cohen, Safety systems based on Image analysis and recognition is not reliable and requires a lot of computer hardware. Insect vision technology represents a departure from traditional image processing technology, based on image processing to recognize objects across varying spatial positions, and distances, in high-scattering mediums. Novel insect vision focuses instead on vision technology that utilizes electromagnetic signals for all-around, fast, high-accuracy multiple threat detection by spectrum signature invariant to their spatial position.

#TeslaWatts and #Qvio®️ answer the question 🤔"Can a #Cybertruck run an AI Agent?"👉 https://lnkd.in/gPev8RtW --- No more digging through owner manuals, endless "how-to" videos, and unhelpful guides—Cybertruck owners could just ask their truck questions and get instant Tesla-approved answers. Now, that's an intelligent driving experience! #View #Ask #Learn #InteractiveVideo #ControlledAnswers #controlledAI #Tesla #DemandFeature

🤖 𝐀𝐦𝐚𝐳𝐨𝐧'𝐬 𝐂𝐥𝐢𝐦𝐚𝐭𝐞 𝐅𝐮𝐧𝐝 𝐁𝐨𝐨𝐬𝐭𝐬 𝐑𝐞𝐜𝐲𝐜𝐥𝐢𝐧𝐠 𝐰𝐢𝐭𝐡 𝐀𝐈 𝐚𝐧𝐝 𝐑𝐨𝐛𝐨𝐭𝐢𝐜𝐬 Amazon and Glacier are piloting technology to reduce landfill waste and increase the use of recycled materials in packaging 👍 Credit: Amazon News ______________ Read TOP AI News and find useful cheat sheets in our Generative AI Daily Newsletter: https://lnkd.in/dH8ag7MA #ai #robotics #artificialintelligence

The potential for business growth through AI is immense, we have all witnessed the exponential growth of NVIDIA, the company that is producing the necessary chip for the AI industry. Today we can witness how AI is revolutionizing almost everything, including recycling processes, by optimizing efficiency and sustainability through various innovative applications. The latest clip from Amazon's Climate Fund only proves that, if ethically or properly used, AI is proving a game changer not only for business, trade, or marketing but for one of the most heated topics in global politics, climate change, or the future of recycling. According to the researcher working on the pilot project, AI algorithms can analyze waste streams, identify different materials, enable automated sorting and separation, as well as increase material recovery all with 99.9% accuracy. Efficiency that couldn't be provided by entire companies in the past. And if this isn't impressive and something to look forward then I don't know what is. #aiadoption #recycling

AI in trucking may have reached an inflection point, where fleets increase their AI adoption and the technologies start living up to their hype. https://bit.ly/3VfEQ4a #trucking #ai #adoption

CNBC's Sarah Min spoke with RJ Assaly about the outlook for #tech and #ai #stocks: "People forget that companies like John Deere are huge in generative AI, right, from computer imaging for selective crop spraying to autonomous tractors to any number of technologies which improve productivity … [Tech is] probably not a big buy at this point, I think the question is, can some of that promise translate into the other 493 stocks in the S&P 500. And, you know, I think it can." See what toggle.ai has to say about John Deere: https://lnkd.in/gKmBbbjU Read the full article on CNBC: https://lnkd.in/g-U7AiMZ

Andrew Brown has such a great point here. In the #trades and any skilled profession, #AI and robots are not going to take over. So long as you keep up with the innovations in your field, with new tools etc. you will have a spot. You may need to adapt, change, and #alwaysbelearning but AI and robots work well only when the humans are helping. We see this all the time with Robo investing. I know some truly believe it is "just as good" or even "better" than an experienced human advisor who may use some of the new AI-driven tools. Still, so far, 100% of the time, I have found serious flaws in the #roboadvisor plans and portfolios I have seen.

The updated Tesla Vision Full Self Driving is based on a single end-to-end AI neural network and replaces over 300k lines of explicit C++ code! To say I’m impressed is an understatement. My mind was blown at how impressive this update is! AI fully in control and navigating end-to-end routes on solely computer vision with no intervention as confidently and safe as a professional driver is wild. Never thought I would live to see the day that, after a few hours driving with AI in control, I’d start to view “manual driving” as something riskier than trusting Tesla. #ai #computervision #selfdriving #tesla