Reply to thread

Message: <blockquote data-quote="kingsultan111" data-source="post: 27088864" data-attributes="member: 575252">I don't have friends at DeepMind. Well in case you missed it or felt bored, read again here then tell me where I'm wrong.People wants to hear what they wants. They don't want to see the reality as with Sri Lanka, millions people voted for a murderer <img src="/styles/default/xenforo/smilies/default/D.gif"  class="smilie" loading="lazy" alt=":D" title="Big grin    :D" data-shortname=":D" />They love fashion, hype but lack the understanding what's really behind hype.  Cocaine and Valley is something I know for decades. I could have been billionaire, but I didn't cause I love honestly. I have no agenda to spend my time here to explain you this.Try again <img src="/styles/default/xenforo/smilies/default/D.gif"  class="smilie" loading="lazy" alt=":D" title="Big grin    :D" data-shortname=":D" />In nut-shell, we live in 1980-1990's with new clothes.> You know what, just try to build a model for a real use case and see how far you can go without understanding underlying implementation. If you think it's just installing the framework and feeding data to it, all you might get to work will be their provided example scripts.You have very narrow mind. I have never said that it's simple as installing framework and feeding data into it. If it's simple as that, the salaries will go down to $10 bucks per year.There are other PhDs apart from Jeremy Howard recommending the top-down non-math approach,<a href="https://machinelearningmastery.com/techniques-to-understand-machine-learning-algorithms-without-the-background-in-mathematics/" target="_blank">https://machinelearningmastery.com/...rithms-without-the-background-in-mathematics/</a>You can use black-box and top down approaches to understand the implementation. Maths is not required for this. You then will be able to select correct algorithms, optimizers, hyper parameters.This will take at least 5 months.None of this require Maths. I have trained an AML investigator with this approach and he was able to built custom models that bought down bank's costs.> Any software that we consider to render some sort of intelligence, we call it AI. Even simple neural networks, that takes some inputs and produces an output based on patterns it has discovered is considered intelligence.Perhaps you have big misunderstanding regarding what we call intelligence. It's not some sort. Intelligence is Intelligence.As Mike <a href="https://spectrum.ieee.org/stop-calling-everything-ai-machinelearning-pioneer-says" target="_blank">https://spectrum.ieee.org/stop-calling-everything-ai-machinelearning-pioneer-says</a> puts here,"They are showing human-level competence in low-level pattern recognition skills, but at the cognitive level they are merely imitating human intelligence, not engaging deeply and creatively"Imitating != Intelligence.Algorithms sometimes makes you feel that it has some sort of intelligence, but in reality they need extensive human training.Also, An intelligence is capable of irrational thought. Most of algorithms you call intelligent has clearly defined sources of input for linear calculation. Human thinking is more associative and can make mental connections for out-of-the-box thinking.Learn more about adversarial attacks against so called "AI" models and you will be able to find how dumb they are.Today, We need to use RL to beat Dota-2, but 80s, when you needed to put a quarter into a machine to play vidya on an 8-bit CPU, the machines had pretty good “AI” which was able to eventually beat even the best players. Google worked really hard to do something which was doable in the 1980s.They basically just reproduced <a href="https://en.wikipedia.org/wiki/TD-Gammon" target="_blank">TD-gammon </a>on a stupid video game.Many people think AI can replace humans. It's not. It's stupid. In late 90's, I remember people thought <a href="https://en.wikipedia.org/wiki/Computer-aided_software_engineering" target="_blank">https://en.wikipedia.org/wiki/Computer-aided_software_engineering</a> would replace programmers.Idiots at Tesla think that they can replace drivers by violating <a href="https://en.wikipedia.org/wiki/Moravec%27s_paradox" target="_blank">https://en.wikipedia.org/wiki/Moravec's_paradox</a>.Today, Auto driving is still stuck at 80's where Mr. <a href="https://en.wikipedia.org/wiki/Ernst_Dickmanns" target="_blank">Ernst</a> driving around the city with Sun-Workstations in his <a href="https://www.politico.eu/article/delf-driving-car-born-1986-ernst-dickmanns-mercedes/" target="_blank">back seat</a>.“AI” doesn’t exist. Machine learning is a force multiplier and productivity enhancer for statisticians. If you believe the “automation”=”no more jobs” crap, machine learning should have at least automated away most statistician jobs. Yet, the statistician (aka “data scientist”) jobs are among the best paid and most in-demand jobs out there at present."AI" hype is created by Silicon Valley oligarchs (well, technocrats) to hide behind algorithms so they can shift liability to algorithms.</blockquote>

Verification: Nawa warak dahaya keeyada? (Namaya wadi kireema dahaya)

Top Bottom