NeoMax’s Post

Google says its new PailiGemma 2 family of models can "analyze images, enabling the AI to generate captions and answer questions about people it 'sees' in photos". Specifically, the model can go beyond simple identification and "describe actions, emotions, and the overall narrative of the scene”. Automatically detecting emotions has been tried for a long time, and claimed to have been achieved as well, but so far "the science stands on shaky empirical ground". Experts are concerned about potential issues with models like this that "can lead to a dystopian future, where your emotions determine if you get the job, a loan, and if you’re admitted to uni”. On a more uplifting and funny side, The Big Bang Theory season 10 episode 14 "The Emotion Detection Automation" has such a machine at its heart. 29 TechCrunch https://lnkd.in/e2R26gEv 𝗚𝗼𝗼𝗴𝗹𝗲 𝘀𝗮𝘆𝘀 𝗶𝘁𝘀 𝗻𝗲𝘄 𝗔𝗜 𝗺𝗼𝗱𝗲𝗹𝘀 𝗰𝗮𝗻 𝗶𝗱𝗲𝗻𝘁𝗶𝗳𝘆 𝗲𝗺𝗼𝘁𝗶𝗼𝗻𝘀 — 𝗮𝗻𝗱 𝘁𝗵𝗮𝘁 𝗵𝗮𝘀 𝗲𝘅𝗽𝗲𝗿𝘁𝘀 𝘄𝗼𝗿𝗿𝗶𝗲𝗱

Time to hold tech giants accountable for unethically using our data. The AI revolution isn't stopping, so how will you adjust your online behavior? We need to steer this tech for our benefit. #AIRevolution #DataPrivacy #TechAccountability

"Decisions made by software should be explainable. When a decision impacts my life, I need to understand what led to it. Perhaps the decision was based on incorrect information. There may be a logical flaw in the decision-making process that I need to question and escalate. Understanding the decision process can help me adjust my actions to achieve better outcomes in the future." Martin's post inspired me to consider where we, as users, would benefit from software's ability to explain algorithmic decisions. For example: Why does my LinkedIn feed show me one post over another? To whom are my posts shown? Why is my booking price different from my friend’s? Why did the flight ticket price suddenly increase? Recently, I had to help an elderly neighbor with a fraudulent banking transaction. We went to the bank, where a young clerk with little experience in transaction fraud took quite a long time to understand the situation and the system's messages and decisions about the transaction. In this era of AI, we should consider ways to make software system decisions easy to explain to everyone. This mindset is more beneficial to people than overregulating AI to protect us from black boxes. What do you think? How important is it for software to explain its decisions?

This resonates with me at so many levels. Starting from rubber duck debugging where I had explained to myself on problems in order to find a solution to giving the context and background of the issue before explaining the decision/recommendation to senior stakeholders. If you are able to explain the "method to the madness", then it is easy for your audience to relate and buy in to your proposal or solution. It has become so much important in the context of AI, but it has always been the case in tech industry. As a side note, I have always appreciated good code comments on why it was done like that (again explainability) versus how it was done, which is self explanatory reading the code. Agree?

As technology is always evolving, it’s essential that we stay on top of new developments to maximize potential benefits and minimize risks. Our Technology Industry Snapshot takes a deep dive into whether children should use generative AI products to complete schoolwork and how US consumers feel about companies that suffer a data security event, closing with our monthly over-index report. Discover tech insights now: https://bit.ly/4e4NBE5

Second episode of our AI blog series, directed by Luka Panic, presents #RAG essentials, addressing its common challenges in order to demystify the technology and ease the learning curve. 🧑💻 You’ll find out how to efficiently use processes of data ingestion and data retrieval while building an application that provides human-like answers about subjects that the LLM isn’t familiar with. 💡 Curious? Read and subscribe ➡️ https://bit.ly/3J5j3VP #PixionAiBlogSeries #LLM

Rather than imposing constraints on AI and algorithms, we should prioritise making their decision-making processes transparent and comprehensible. By fostering explainability, we can build trust, identify biases, and ensure that these powerful tools are used responsibly and ethically. This approach not only mitigates risks but also unlocks the full potential of AI to benefit society.

Announcing an unvetted, unbounded, and incomplete AI model that you then choose to 'pause' citing safety concerns gets you a lot of publicity on the initial announcement ("OMG the world has changed!", "Holy crap that's outrageous how dare they!!) and a lot of publicity on the subsequent backtrack ("See, it's so powerful they have to contain it!!!"). Or perhaps it was never ready and was announced as part of a wanton hype machine that deploys vivid impressions to buy time to actually build other more boring stuff in the background that rides on your massive PR ... and gives you space and public expectation to build the next wanton vividness, softening the market up for accepting what two months ago would have been unthinkable. This is not a new business model, and is being repeated again here because it is so successful.

Google DeepMind Introduces SIMA: The First Generalist Artificial Intelligence AI Agent to Follow Natural-Language Instructions in a Broad Range of 3D Virtual Environments and Video Games The pursuit of artificial intelligence that can navigate and comprehend the intricacies of three-dimensional environments with the ease and adaptability of humans has long been a frontier in technology. At the heart of this exploration is the ambiti... https://lnkd.in/ekS_5Dku #AI #ML #Automation

The next era of Gemini! Big AI news from Google DeepMind! Introducing Gemini 2.0, the most advanced multimodal AI model yet. With cutting-edge features for spatial reasoning, video understanding, and real-time interactivity, Gemini 2.0 paves the way for the "agentic era." Developers and businesses can now leverage its enhanced APIs to create dynamic, future-ready applications. Source: https://deepmind.google/ #ArtificialIntelligence #DeepMind #Innovation #AI4Business #TechTrends