Elvis D.’s Post

𝐄𝐱𝐩𝐞𝐫𝐢𝐦𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧 𝐅𝐎𝐌𝐎 Should you be more data-driven and experiment more? Amazon Science teamed up with a Nobel-prize-winning economist to help us answer that question. Paper: https://lnkd.in/gF_cSa6G (cc James McQueen) 𝐖𝐡𝐲 𝐢𝐭 𝐦𝐚𝐭𝐭𝐞𝐫𝐬: Big tech companies have opened an innovation gap over the competition. Amazon, Google, etc. run hundreds of thousands of experiments yearly. 𝐃𝐢𝐯𝐞 𝐝𝐞𝐞𝐩𝐞𝐫:  The paper develops a methodology to measure the value of a new experiment. If the value is high, this means you should run more experiments.  The paper uses an empirical Bayes decision framework to estimate: 1. A prior: what you expect the impact of your launch to be, without running an additional experiment 2. A posterior: the prior updated with what you learn from a new experiment Then: The value of a new experiment = (1) The benefit from learning (better launch decisions) - (2) the cost of running it. 

Hello everyone, 💡 **Embracing Failure: A Step Towards Growth** 💡 I recently applied for the Amazon Summer ML School and, unfortunately, I wasn't shortlisted. While it’s disappointing, I believe that failure is an essential part of the learning process. This experience has taught me the importance of resilience and perseverance. It has motivated me to work harder, improve my skills, and prepare even better for future opportunities. I believe life is a zigzag path with ups and downs,but everything happens for a reason. Failures are not setbacks but rather stepping stones to success. I'm committed to learning and growing, and I’m excited to see where this journey takes me next.As Thomas Edison once said "I have not failed I've just found 10,000 ways that won't work". Let's celebrate our journeys,both the highs and lows. #FailureIsPartOfSuccess #GrowthMindset #MachineLearning #Resilience #LearningJourney # stay positive.

As we are about to start 2024, I will like to share this piece by Werner Vogels Amazon's chief technology officer, strategy for taking effective notes. “... one of the worst feelings I can think of, [is] to have had a wonderful moment or insight or vision or phrase, to know you had it, then lose it.” Anne Lamott Also an old Chinese proverb says that “the faintest ink is more powerful than the strongest memory.” Advice: Take notes, Lots of them. https://lnkd.in/duGe_avC #Amazon #AWSreinvent #AWSemergingtalent

Hi Everyone, Why is only success celebrated on LinkedIn? Failure is also a good teacher. There are always two sides: one that people want to see and another that they don't, but both are part of our journey. I believe life is a zigzag path with ups and downs, but everything happens for a reason. I recently applied for the Amazon ML Summer School 2024 but wasn't shortlisted. While it’s easy to only share our successes, it's important to acknowledge and learn from our failures too. As Thomas Edison once said, "I have not failed. I’ve just found 10,000 ways that won’t work." A big thank you for making me aware of this opportunity and to Amazon for providing such valuable learning experiences. I am hopeful for the best in the future and excited to keep learning and growing in the field of machine learning. Let's celebrate our journeys, both the highs and the lows. Amazon #MachineLearning #FailureIsAStepToSuccess #LearningJourney #AmazonML #vitblions

The Sri Lankan who solved a maths problem for Jeff Bezos & gave the world Amazon! Read below to know the entire story. Bezos animatedly described an anecdote from their days at the university, when he and a roommate failed to solve a math problem despite wracking their heads over it for hours. He was then an aspiring theoretical physicist. When they couldn’t, Bezos, both he and his roommate screamed in unison “Yasantha”, because they knew the Sri Lankan would help them out. When they went to his room, according to Bezos, Yasantha stared at the problem for a while and then gave them the answer. Taken aback by the promptness, Bezos asked him, “Did you just do that in your head?” “No, that would be impossible,” Yasantha said, adding that he only knew the answer because he had solved a “very similar problem three years ago”. “I was able to map this problem on to that problem, and the answer was immediately obvious…” he added. “That was an important moment for me,” the Amazon founder said, “because that was the very moment I realised I was never going to be a great theoretical physicist.” The two haven’t kept in touch since 1985, but clearly share a mutual admiration. “Jeff was an excellent student, and a very persistent, tenacious one,” said Yasantha, “That is unique to him.” Yasantha described how students once dared each other to complete a computer science assignment in a single line of coding. “Finally, I gave up and did this in 10 or so lines of code,” he added, “But I remember Jeff working through all night, in pursuit of the most compact solution, and turned in a two-line solution that was probably the shortest anyone could do…” There would have been no Amazon if Jeff had solved that numerical and had pursued Theoretical Physics. Incredible story❤️ __ Follow Arin Verma Source: In comments #india #business #startup #amazon #maths Subscribe to my newsletter 'BusinessBuzz.' Link is in comments.

“Something more emergent and collaborative is seen as an anti-pattern” Sounds right! I think this has less to do with “big” company culture and more to do with the fact that cooperative folks with experience have gotten burnt by less cooperative folks at least a few times in their careers, so they learn that being less cooperative in the short term (less than 5 years) is acceptable and existentially required. Without deliberate alignment and incentives for cooperation silos develop naturally and organically out of self interest…. PRDs remove the need for individual contributors (not on the product team) to be anything more than order takers who have to deliver on requirements by a certain deadline instead of take responsibility for the outcomes of the product itself and absolves the organization to learn what features actually achieve business results. As a result, innovation suffers and many opportunities are passed over…leaving the fate of the company in the hands of the CEO who is usually too far removed from the problems at hand.

🌟 Embracing Setbacks as Stepping Stones 🌟 Success stories often shine on LinkedIn, but acknowledging setbacks is equally important. Though I wasn't selected for the Amazon ML Summer School, I'm more determined than ever to advance in machine learning. This experience fuels my commitment to continuous learning and growth. Congratulations to all who made it! Your achievements are truly inspiring. Let's keep pushing the boundaries of innovation together. Onward and upward! 🚀 #Amazon #MachineLearning #ML #GrowthMindset #Resilience #ContinuousLearning

What are some plausible reasons why big tech holds on to traditional PRDs as a practice? I've been thinking about this for a while, and have a couple ideas: 1. They are "big" companies. As much as we like to think of tech companies as nimble and embracing new practices, they are still big companies that hold on to their habits. 2. The engineering culture. In a lot of big tech there are strong incentives for functions to stay in their lane. Product *should* X. Engineering *should* Y. Design *should* Z. And if everyone does their thing, they get promotions. Engineering time is seen as precious, and PRDs are a signal that product is doing their thing. Something more emergent and collaborative is seen as an anti-pattern. 3. Related to #2, they form the documentation chain for promotions. In a lot of big tech, there is a strong demand to demonstrate your impact. Documents and a document trail satisfy that need. 4. Related to #2 and #3, the teams can be more work groups than actual teams. Everyone wants their high profile project. Engineering leads negotiate with PMs to see who gets to do what. These PRDs play into that. 5. Less cross-functional collaboration than you might expect. This may be related to #2. "I've had engineers tell me to go away and not come back until I have a PRD. And design is off in their own world somehow." 6. Dependencies. A dirty secret is that outside Amazon (and frankly even in Amazon) there are way more dependencies than people would like to accept. This means that there is a strong need for premature convergence. 7. Understaffed legal teams that want something to review. These are just off the top of my mind talking to friends in these companies. What else? (Oh, I'm sure someone will ask what the alternative is, and the answer is basically something more emergent, more participatory design and writing, less premature convergence on a "problem definition", etc.)

Big news from Amazon Web Services. They're investing a whopping $10 billion in this state. This isn't just any investment; it's the largest in the state's history, four times bigger than any before. Subscribe to our tech newsletter, the Blueprint which is your daily source of tech, science and engineering innovation.  Here you go: 👉 https://ie.social/V0ymX #engineering #science #technology