<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Minhaaj Rehman on Medium]]></title>
        <description><![CDATA[Stories by Minhaaj Rehman on Medium]]></description>
        <link>https://medium.com/@minhaaj?source=rss-95f3a1e37177------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*PI5rcLmC_Gl2mK1jVcr-Mw.jpeg</url>
            <title>Stories by Minhaaj Rehman on Medium</title>
            <link>https://medium.com/@minhaaj?source=rss-95f3a1e37177------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 22:31:28 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@minhaaj/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Annoucement — Moving to Substack]]></title>
            <link>https://medium.com/@minhaaj/annoucement-moving-to-substack-98e0ed2aad27?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/98e0ed2aad27</guid>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Thu, 25 May 2023 12:09:04 GMT</pubDate>
            <atom:updated>2023-05-25T12:09:04.121Z</atom:updated>
            <content:encoded><![CDATA[<h3>Annoucement — Moving to Substack</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/964/1*3hs5euDAT_-w1hG3mH_zKw.jpeg" /></figure><p>After careful consideration and upon request from readers, i have decided to move my blog to Substack. You can follow my work at <a href="http://minhaaj.substack.com/">minhaaj.substack.com</a></p><p>The reason being that i am able to provide better content with community features. For example, for premium articles i am able to provide more rich experience and offer subscriber chats. You can pledge your support via a small fee and can access to full length in-depth articles, trends, podcasts and much more. In subscriber chats you can ask questions and book one-to-one meetings where i provide mentorship and advisory. Podcast sponsors and advertisers can also reach out to let me know about your products/services. Biggest of all reasons is that you will get all the news delivered to your inbox right after its published. On Medium there is a chance its lost in the scrolling. You can also reach out to ask for my Notion productivity templates, my books and infographic services from my company — Psyda.</p><p>If you are a non-profit that needs help, feel free to drop me a line and we’ll see how we can help you out. Psyda does a lot of philanthropic work across the world and we’d love to help you out. You can head over to psyda.org to find out more about us.</p><p>I also have a buymeacoffee account for those who are doing the noble work by supporting me and keeping everything ad-free. Patrons like you are greatly appreciated. You can help me out by one-time donation or making a pledge for monthly support. Heres the QR code for my page.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/1*ektoNXq3hWK9iIX-3s9ySQ.png" /></figure><p>Lastly, the most important bit of all is staying curious and asking questions. We live in a time where boundaries of facts and fiction have blurred and disinformation and hype has taken over profundity. Let me know the topics you’d like me to cover and give feedback on what you like and what you don’t. I like to hear opposing views and opinions and learn from others. This is the essence of erudition in my opinion and rising above your beliefs is what makes us the enlightened beings.</p><p>Thanks again for your support and love over the years!</p><p>See you on substack!</p><p>Regards,</p><p>Minhaaj</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=98e0ed2aad27" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[US Congress hearing and Monopoly in AI. The ChessBoard]]></title>
            <link>https://medium.com/@minhaaj/us-congress-hearing-and-monopoly-in-ai-the-chessboard-1e7e4aa95017?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/1e7e4aa95017</guid>
            <category><![CDATA[sam-altman]]></category>
            <category><![CDATA[microsoft]]></category>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[congress]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Fri, 19 May 2023 16:44:27 GMT</pubDate>
            <atom:updated>2023-05-19T16:44:27.830Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*bXYLYVZEmTnQibe7.png" /></figure><p>The hearing before the Senate Judiciary Subcommittee on Privacy, Technology and the Law offered congressional members the chance to question Sam Altman and other tech leaders about the “urgent” need to create regulations around AI.</p><p>Senators questioned Altman, Gary Marcus, and Christina Montgomery the chief privacy and trust officer at IBM, about the need to AI regulations. I still wonder why Gary Marcus though. Ironically Sam didn’t sign Elon’s letter some time ago.</p><p>Suddenly his skin has thinned and he wants an agency to regulate AI beyond a certain threshold of AI capabilities and license further development, preferably post-dated as Microsoft should permanently be exempted from such laws. IBM a failing company is joining the lead however sees no support from the rest of the tech giants. Is it also a surprise that Microsoft recently laid off its ethics team recently. Why the crocodile tears now?</p><p>Devil is in the details. The guy doesn’t want self-replicating models which gives out his monopolistic mindset. I think he was pretty young to remember the 1999 lawsuit against Microsoft that split it the last time. As ridiculous as the whole event is, Gary Marcus supported global standards as if the US has any power over them. After betraying the open source community by taking their work and data, and calling for oligarchical control of the politics it is a new low.</p><p>This scapegoat who works for health insurance and has no equity in OpenAI doesn’t realize that once Microsoft owns the shares he’s just a mouthpiece. All he has to do is to see DeepMind’s attempt to get rid of Google. That turned out very well. IBM just laid off 7800 employees thanks to AI now wanting regulation. Do you smell something brewing or they’ve taken people for utter fools?</p><p>Here are some global domination agendas of these sellouts. Complete domination of public data and the content, amount, and timing of its release to the public in a manner that suits them. Acquiring new data by mere interaction of people with it while identifying their demographics and usage pattern or further use without consent. Global dissemination of Western propaganda and pushing occidental narratives to infiltrate the people with differences of opinions. The regulation also means persecuting people who voice their concerns against such schemes by demonetization and de-platforming as US media companies have done for decades.</p><p>This is how Noam Chomsky’s proverbial manufacturing the consent will happen. The public in their hype frenzy will hand over their freedoms to find the solution to the problems they self-created and play along giving legitimacy to their crimes in the name of progress. A faltering US government will pass any laws to gain donations and support for upcoming elections.</p><p>Its then you’ll cede unilateral control of what’s fiction and what’s reality and there will be a monopoly on what’s information and what’s a lie, Choose wisely.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1e7e4aa95017" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[UMC & TSMC semiconductor rivalry in Taiwan]]></title>
            <link>https://medium.com/@minhaaj/umc-tsmc-semiconductor-rivalry-in-taiwan-fd6f8ac02d64?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/fd6f8ac02d64</guid>
            <category><![CDATA[semiconductors]]></category>
            <category><![CDATA[china]]></category>
            <category><![CDATA[taiwan]]></category>
            <category><![CDATA[tsmc]]></category>
            <category><![CDATA[umc]]></category>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Tue, 16 May 2023 15:39:31 GMT</pubDate>
            <atom:updated>2023-05-16T15:39:31.582Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sKW1TpbNEvR97LNxagKgiQ.jpeg" /></figure><p>First mover disadvantage in the semiconductor race<br>============================================</p><p>In the 1970s, Taiwan faced new economic challenges. Its single biggest export category was textiles. But import quotas and international competition limited that industry’s future.<br>Taiwan’s electronics industry was then its second-largest export category. However, most of that industry still consisted of low-value electronics assembly operations.</p><p>It was, however, in the genius of clairvoyant Taiwanese back in the 70s, that the semiconductor supremacy rests. The saga of a great rivalry between the pioneer of the semiconductor industry UMC and the following TSMC is a riveting case study of a few smart business decisions.</p><p>Various Taiwanese who spent time in the United States pushed for Taiwan’s economy to pivot towards integrated circuits.<br>The government studied previous technology transfer situations and why they failed. They determined that the most common reasons for failure were insufficient R&amp;D investment, lack of peripheral support after the transfer, and untrained personnel.</p><p>So the government set up the Industrial Technology Research Institute or ITRI to thoroughly acquire, introduce and transfer new technologies to the private industry. They would oversee the whole process like a project manager.</p><p>In late 1979, ITRI set up a preparatory office for what would eventually be UMC — Taiwan’s premier vertically-integrated electronics manufacturer. An Integrated Design Company, so to speak.</p><p>ITRI even selected the products that UMC would sell. They transferred IC design IP for ten products, including telephones, toys, timers, calculators, and watches. Mostly items that the Taiwan electronic industry itself exported.</p><p>The Taiwanese government then reached out to several large Taiwanese companies to get them to invest in UMC. Many of Taiwan’s biggest companies including (ironically) China Development Financial, TECO, and the Taiwanese Bank of Communications contributed $12.5 million.</p><p>TSMC and UMC have long been rivals. This dates back to Tsao’s “accusation” that Morris Chang stole the independent foundry idea, and then prevented UMC from adopting the model until after TSMC was founded.</p><p>This idea — which Tsao later tweaked — sounds similar to although not quite entirely like the independent foundry idea that Morris Chang and TSMC pioneered.</p><p>The main difference was that Tsao’s proposal would only serve design houses that UMC invested in — mostly favoring those in Taiwan. Chang’s dedicated foundry would take on all comers.</p><p>Carver Mead — the Caltech professor famous for revolutionizing VLSI design — proposed the concept of a “fabrication facility” all the way back in 1979. He called it a “silicon press”.</p><p>From about 1993 onwards, UMC managed to keep up with TSMC. Then it made a colossal mistake of partnering with IBM on copper interconnect technology while TSMC rejected it. A fatal mistake!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fd6f8ac02d64" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ScyllaDB and Discord. A migration Success Story with Trillions of Messages.]]></title>
            <link>https://medium.com/@minhaaj/scylladb-and-discord-a-migration-success-story-with-trillions-of-messages-455b2d3bb45?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/455b2d3bb45</guid>
            <category><![CDATA[scylladb]]></category>
            <category><![CDATA[database]]></category>
            <category><![CDATA[data-engineering]]></category>
            <category><![CDATA[cassandra]]></category>
            <category><![CDATA[discord]]></category>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Tue, 09 May 2023 18:33:34 GMT</pubDate>
            <atom:updated>2023-05-09T18:33:34.425Z</atom:updated>
            <content:encoded><![CDATA[<h3>ScyllaDB and Discord. A Migration Success Story with Trillions of Messages.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MK8L9_Un_nCEQmV6kEHSJQ.jpeg" /></figure><p>Seven years ago <a href="https://www.linkedin.com/in/ACoAAAA7qnUBsqm-7t7OZUZ9lYncioSLZDx4DDA">Jason Citron</a> and <a href="https://www.linkedin.com/in/ACoAAAKMejcBtvzYXdKOfonvGlW_RB-0f2SNdIM">Stanislav Vishnevskiy</a> joined hands together to create a platform, that the world would remember as the future of the social communication community, <a href="https://www.linkedin.com/company/discord/">Discord</a>. Discord runs on Windows, macOS, Android, iOS, iPadOS, Linux, and in web browsers. As of 2021, the service has over 350 million registered users and over 150 million monthly active users.</p><p>This isn’t for the faint of the heart though because storing and managing a trillion messages per day is not a joke. However, this is where their remarkable journey into data architecture started. They started from <a href="https://www.linkedin.com/company/mongodbinc/">MongoDB</a>, and migrated to Cassandra hoping that it’ll put their miseries to an end but that was just the beginning. At the beginning of 2022, it had 177 nodes with trillions of messages and things were getting out of hands.</p><p>In Cassandra, reads are more expensive than writes. Writes are appended to a commit log and written to an in-memory structure called a memtable that is eventually flushed to disk. Reads, however, need to query the memtable and potentially multiple SSTables (on-disk files), a more expensive operation. Lots of concurrent reads as users interact with servers can hotspot a partition, which they refer to imaginatively as a “hot partition”. The size of their dataset when combined with these access patterns led to struggles for their cluster. This also meant that this architecture had a catastrophic flaw that servers with 2 users would get as many resources as huge servers like Midjourney.</p><p>Enters <a href="https://www.linkedin.com/company/scylladb/">ScyllaDB</a>, a Cassandra-compatible database written in C++. Its promise of better performance, faster repairs, stronger workload isolation via its shard-per-core architecture, and a garbage collection-free life sounded quite appealing. To write a migration tool they wrote it from scratch in Rust because using existing tools the migration time for that huge data was 3 months.</p><p>Rust touts fearless concurrency as one of its main benefits — the language should make it easy to write safe, concurrent code. Its libraries also were a great match for what they were intending to accomplish. The Tokio ecosystem is a tremendous foundation for building a system on asynchronous I/O, and the language has driver support for both Cassandra and ScyllaDB. And by the way migration was done in 9 days! 3.2 million messages per second. Let that sink in. Not like Elon Musks’ sink!</p><p>Whats the difference?</p><p>They came down from 177 cassandra nodes to 72 ScyllaDB nodes<br>Latency to fetch went down from 40–125ms to 15 ms<br>Message insert went from 5–70ms to consistent 5 ms.</p><p>What’s the lesson? That as much as data engineers love to play with novel architectures and simply add nodes to avoid work, thoughtful research and brave initiative to implement changes is what matters in the day. I bet they are saving millions of dollars just by those changes. One of the many success stories of engineering success!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=455b2d3bb45" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Geoffery Hinton’s Resign and LangChain Framework]]></title>
            <link>https://medium.com/@minhaaj/geoffery-hintons-resign-and-langchain-framework-217aaf115bab?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/217aaf115bab</guid>
            <category><![CDATA[python]]></category>
            <category><![CDATA[langchain]]></category>
            <category><![CDATA[geoferry-hinton]]></category>
            <category><![CDATA[google]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Thu, 04 May 2023 16:13:45 GMT</pubDate>
            <atom:updated>2023-05-04T16:13:45.023Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RF-ehHhHDBc9pOGpmz7Ksw.jpeg" /></figure><p>As Geoffery Hinton resigns from Google finally after a decade-long gig citing ethical reasons, there is more to surmise why now? Michio Kaku calls ChatGPT computer-based plagiarism and LLMs a hodge podge of truth and falsehood. Why ardent supporters of AI are now waking up to its collateral damage?</p><p>In my opinion, it has less to do with the technology itself which is flawed, unaudited, and overly glorified owing to its paid positioning but with what can be built and sold from its promise. One tool that is taking this promise to a whole new level and is gaining rapid popularity amongst developers is LangChain. Released last year in October by Harrison Chase it already has attracted a remarkable set of core contributors and creators and is ranking among the top frameworks in AI. They recently announced a $10 million seed round.</p><p>There are two main value props the LangChain framework provides:</p><p>Components: LangChain provides modular abstractions for the components necessary to work with language models. LangChain also has collections of implementations for all these abstractions. The components are designed to be easy to use, regardless of whether you are using the rest of the LangChain framework or not.</p><p>Use-Case Specific Chains: Chains can be thought of as assembling these components in particular ways in order to best accomplish a particular use case. These are intended to be a higher-level interface through which people can easily get started with a specific use case. These chains are also designed to be customizable.</p><p>It has six distinct components:</p><p>Schemas<br>Models<br>Prompts<br>Indexes<br>Memory<br>Chains<br>Agents</p><p>You can use multiple models from OpenAI, Cohere, and Hugginface into the framework and pipe that into subsequent modeling and prompts. Its excellence, however, is its ability to connect nuanced data sources like Wikipedia, google, Reddit, quora, and practically any API-based data source and post-process and chain it in a series of prompts and actions to create endless workflow ideas.</p><p>As an example, it can run a whole travel agency on its own given all datastreams are available to its engine including voip-based calling to confirm and update changes. It can access weather API to update travelers, inform hotel managers of arrival schedules of guests and delays as well as medical record-based alerts to destination hospitals. Its just the tip of the iceberg. In the latest video by <a href="https://www.linkedin.com/in/ACoAAADJqKcBW1qQmvqN4ow6F2YDxD5nAjHa1oE">Joanna Stern</a>, she was able to buy pass her bank security by using her AI-generated voice.</p><p>Eventually, people can almost create a digital extension of themselves that does the gruntwork for them including some very complicated and artistic work. While EU as always is the sane head in regulation, the US is spearheading destruction bypassing all laws. Writers in Hollywood just demonstrated against ChatGPT’s stealing of their proprietary text and commercializing it to cut out their jobs. What’s next?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=217aaf115bab" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cohere’s AI powered semantic search]]></title>
            <link>https://medium.com/@minhaaj/coheres-ai-powered-semantic-search-9bae2ac72e48?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/9bae2ac72e48</guid>
            <category><![CDATA[search]]></category>
            <category><![CDATA[semantics]]></category>
            <category><![CDATA[cohere]]></category>
            <category><![CDATA[google]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Tue, 02 May 2023 15:06:41 GMT</pubDate>
            <atom:updated>2023-05-02T15:06:55.398Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yZr7QDhVnEq8o5MV3d5mxw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iTOv4ynXecq0a22RQ33tXQ.png" /></figure><p>Google remains the dominant search engine for decades now despite its sub-par results and irrelevant crawling. This dominance is finally coming to an end with search powered by AI that isn’t its usual keyword-based search.</p><p>You see when you search for ‘Persian cats’ in google, your options are either ‘search’ that will bring you all the paraphernalia from the internet to scour through or ‘I am feeling lucky’ if you are confident that your search is pedestrian.</p><p>What it does is that it matches your keywords with the highest-ranking sources with relevant keyword displays regardless of their semantic meanings and linguistic coherence. Now what if generative models and LLMs provide a better way to eliminate this manual dumpster diving for gems from 36,000,000 results? Unless you are intentionally searching for web pages, you want a well summarize version of the topic you searched for.</p><p>In contrast, a <a href="https://www.linkedin.com/company/cohere-ai/">Cohere</a>’s semantic-based search system can contextualize the meaning of a user’s query beyond keyword relevance, allowing it to return more relevant and accurate results.</p><p>The Cohere Rerank endpoint is designed to bridge this gap. And what’s more, Rerank delivers much higher quality results than embedding-based search, and it requires only a single line of code change in your application.</p><p>Adding Rerank to your search stack is easy. Once you retrieve the initial results from your existing search engine, pass the initial query and list of results into the endpoint like so:</p><p>results = co.rerank(query=query, documents=documents, top_n=3, model=”rerank-multilingual-v2.0&quot;)</p><p>Here are what the arguments represent:</p><p>query: the user query text<br>documents: list of candidate results you want to rerank<br>top_n: the number of reranked results to return</p><p>As the results show, lexical search can find relevant search results, on average, for about 44% of the search queries. Hence, for the majority of search queries, the user doesn’t get any relevant information in the top three results. Embedding-based semantic search can boost this to 65%, while Rerank achieves the best performance: for about 72% of search queries, we were able to find and show the most relevant hit among the first three results.</p><p>Rerank can also be used with embedding-based semantic search, which will yield even better results.</p><p>Its still in process but the promise is huge and groundbreaking. Arguably the next best thing happening to knowledge management. (blogpost link in comments)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9bae2ac72e48" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[McKinsey’s MECE & OpenAI]]></title>
            <link>https://medium.com/@minhaaj/mckinseys-mece-openai-c7a07ba4eb38?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/c7a07ba4eb38</guid>
            <category><![CDATA[deepmind]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[managementconsulting-firm]]></category>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[mckinsey]]></category>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Fri, 28 Apr 2023 13:35:32 GMT</pubDate>
            <atom:updated>2023-04-28T13:35:32.014Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*-CqFD_tTWTHlmN7OLHojIA.jpeg" /></figure><p>In 2001, Ethan Rasiel wrote his classical book The McKinsey Mind, narrating his lessons from interviewing <a href="https://www.linkedin.com/company/mckinsey/">McKinsey &amp; Company</a>, a strategic consulting firm par excellence. For many, it serves as the bible of management consulting. As business school students this is recommended to every aspiring grad student.</p><p>While Mckinsey for the most part attracts brilliant talent and executes its projects better than most, there are times it forgets its own lessons. Going after big projects blurs the vision and principles of many organizations and McK is no exception like its historical fiasco at CNN+. McKinsey originally expected CNN+ to bring in around 2 million subscribers in the U.S. in the service’s first year and 15–18 million after four years. They counted every individual airport tv as a subscriber. Or its $600 million settlement with drug companies for its shameful role in normalizing opioids.</p><p>Jeff Skilling, CEO of Enron was groomed from this nursery of finely educated crooks who is serving 24 years in prison for a federal felony related to arguably one of the biggest money heists in history.</p><p>One concept of MECE (pronounced “mee-see” and an acronym for Mutually Exclusive, Collectively Exhaustive), is a basic tenet of the McKinsey thought process. Being MECE in the context of problem-solving means separating your problem into distinct, nonoverlapping issues while making sure that no issues relevant to your problem have been overlooked.</p><p>Nonetheless, the book is very focused on a system of elaborate knowledge management that has added a lot to the body of literature on business management. I often get asked how I deal with complex, multifaceted, and multifarious data in organizations I work with and I think some of the playbook rules from this book are worth quoting:</p><p>* Find the key drivers.<br>* Look at the big picture.<br>* Don’t boil the ocean.<br>* Sometimes you have to let the solution come to you.</p><p>Let’s apply these rules to separate the chaff from the grain. A company’s performance is the sum total of its Income minus expenses and other non-tangibles. What drives income is a product/service that brings in revenue. Now take a company like OpenAI and DeepMind and see their balance sheets to the extent they are available and a blind man can read how they are bad businesses.</p><p>This brings us to the bigger picture which is a promise that for $20 a month, you can plagiarize your homework, pass medical exams, and clear LSATs to become a lawyer. A class lawsuit with some decent lawyers can easily take that promise and revenue away easily.</p><p>Boiling the ocean means doing more than necessary to get results and on that count its right on track to doing nothing but perpetuating falsehood as science and paying off journalists to write fiction.</p><p>Next time, you are being sold AI without numbers, remember CNN+ and MECE.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c7a07ba4eb38" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[China’s AI supermacy starts from NEC in Silicon Valley]]></title>
            <link>https://medium.com/@minhaaj/chinas-ai-supermacy-starts-from-nec-in-silicon-valley-3d1250f81059?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/3d1250f81059</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[deep-learning]]></category>
            <category><![CDATA[china]]></category>
            <category><![CDATA[baidu]]></category>
            <category><![CDATA[nec]]></category>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Wed, 26 Apr 2023 14:42:21 GMT</pubDate>
            <atom:updated>2023-04-26T14:42:21.018Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rBY1lKtvG3o7LtgrFS8tuw.jpeg" /></figure><p>This week I intend to write on the humble beginnings of Chinese AI and its ultimate triumph over the vicissitudes of deep learning and regulatory buy-in. Through years of grit, perfectionism, and hard work, China became the undoubted leader in AI, not without its woes.</p><p>China’s AI development back decades to a crucial wellspring: the Silicon Valley lab of NEC (a Japanese multinational information technology and electronics corporation).</p><p>Last week <a href="https://www.linkedin.com/in/ACoAAA8Pk38ByNewz4vwDhl--LLW-yGb_Istksg">Jeffrey Ding</a>’s phenomenal substack publication China AI News published a two-part piece on how China’s current AI success traces its roots in NEC’s state-of-the-art deep learning lab. After Google fully withdrew from the Chinese mainland market, Baidu was flushed with success. Robin Li, who is sensitive to technology, kept up with the dynamics of top technology teams at home and abroad, and soon realized that AI would bring new opportunities to Baidu, and began to look for talents everywhere.</p><p>When the two met, Yu Kai preached on deep learning, and Robin Li talked about his interest in speech recognition and image recognition. There was the surging excitement of two people who regretted not having met earlier. It was also on this day that Yu Kai and Robin Li made a historic mutual choice for Baidu and China’s AI landscape: Robin Li invited Yu Kai to return to China. It was only after Yu Kai joined that Baidu took the lead in injecting deep learning genes.</p><p><a href="https://www.linkedin.com/in/ACoAAAACAQgBdBZZ2ajAJYPtqAfRvF-OiG_N2oI">Robin Li</a> prepared to establish the world’s first AI laboratory named after “deep learning” — the Baidu Institute of Deep Learning (IDL).</p><p>Yu Kai was the actual operator of Baidu IDL in the early days. Under his influence, Baidu attracted a large number of outstanding AI scientists, such as <a href="https://www.linkedin.com/in/ACoAAACslaIBLHsFoEnCV0RN45QRc83vFzOGPG4">Wei Xu</a>, the founder of the PaddlePaddle platform, <a href="https://www.linkedin.com/in/ACoAAAnQfqgBezOo4JQ_MAbL2l2P40XvmN7Pge8">Chang Huang</a>, the co-founder of Horizon, and <a href="https://www.linkedin.com/in/ACoAAAGzSc4BUY5MFLd55DJ0phqC5IFYfsGZroI">Yuanqing Lin</a>, the dean of Baidu Research Institute. Coincidentally, they have all visited the same place in the past: Silicon Valley’s NEC Lab.</p><p><a href="https://www.linkedin.com/in/ACoAAAABD2EBzvx-hcsZWQGyS4ZHJqULlscaUdk">Yihong Gong</a>, introduced the first group of Chinese AI researchers to NEC, including Wei Xu and Shenghuo Zhu who interned at NEC while completing a doctorate in machine learning at the University of Rochester.</p><p>When the Chongqing Research Institute of the Chinese Academy of Sciences was being established in 2011, the dean Jiahu Yuan went three times to the U.S. to invite Zhou to join, so he decided to return to China to make contributions and attract talents for the country. Four years later, he further implemented face recognition technology and founded Cloudwalk.” Cloudwalk is one of China’s top computer vision companies.</p><p>For translation of the original Leiphone article published in Chinese, The past happenings of NEC Lab in Silicon Valley: the one who dragged Chinese companies into the AI era, see comments.</p><p>And they did it without shooting down childish balloons, restricting semiconductors, and questioning social media apps. Thats class. Success is the best revenge.</p><p><a href="https://chinai.substack.com/p/chinai-221-nec-silicon-valley-as">https://chinai.substack.com/p/chinai-221-nec-silicon-valley-as</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3d1250f81059" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Spotify Algorithm and Product-Market Fit]]></title>
            <link>https://medium.com/@minhaaj/spotify-algorithm-and-product-market-fit-3b453c352905?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/3b453c352905</guid>
            <category><![CDATA[spotify]]></category>
            <category><![CDATA[music]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[wsj]]></category>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Wed, 19 Apr 2023 14:22:26 GMT</pubDate>
            <atom:updated>2023-04-19T14:22:26.089Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*YSaozwU5zzR3OVPMdq4N1A.png" /></figure><p>A couple of days ago, <a href="https://www.linkedin.com/company/the-wall-street-journal/">The Wall Street Journal</a> published a deep-dive piece on <a href="https://www.linkedin.com/company/spotify/">Spotify</a>’s algorithms that it uses for recommending songs and playlists to its users. In the past, I’ve written about the bitter-sweet experiences with Spotify’s music and its evolution over the years.</p><p>Spotify was built upon the shoulders of giants like Pandora and <a href="http://last.fm/">Last.fm</a> with a recipe of its own. As one of the invite-only members of it in 2009, I can vouch for the creativity it used to make recommendations that were far better than its competitors intertwined with its minimalist Scandinavian UX. The almost instantaneous sub-100ms query times and machine learning algorithms it used are what gave it the impetus in a rather saturated market back then. It was based on a very brave decision like the money ball guy. It was to acquire an analytics company called Echo Nest that used machine learning to recommend music.</p><p>The Echo Nest began as a research spin-off from the MIT Media Lab to understand the audio and textual content of recorded music. Its creators intended it to perform music identification, recommendation, playlist creation, audio fingerprinting, and analysis for consumers and developers. The Echo Nest was founded in 2005 from the dissertation work of Tristan Jehan and Brian Whitman at the MIT Media Lab.</p><p>It initially used a technique called collaborative filtering recommending music based on public and private playlists and the Euclidean distance between two songs and the probability that listeners would like more similar items. As expected, human beings aren’t the most logical of organizers and their mood swings and seasonality isn’t good for algorithms trained on that flawed data. As a remedy, what they did was that they added an additional layer of recommendation algorithm based on metadata and embeddings from the sound characteristics like loudness, bass, temp and hell ya, danceability.</p><p>All is well and good so far but here comes the algorithmic bias problem. You see the algorithm trained on Western music data recommends some horrible results in Eastern markets. Imagine data trained on Taylor Swift’s songs recommending songs to Indian audiences. Let’s leave it at an unsavory shudder for now. On the other hand, if these markets don’t have enough songs for a specific genre you have a cold start problem on top of a class imbalance misery (pun intended).</p><p>Wish AI competence was all a company needed. This year it shutdown its Spotify Live app after two years of experimentation. Earlier this year Spotify Removes Indian Label Zee Music Company’s Catalog in Licensing Dispute. This has happened more than once. Record label and Copyrights around music are unrelenting. Earlier it lost its chance to be Youtube while it had it. Its shares are about the same price they were in 2018. It has 500 million customers that they don’t know how to monetize.</p><p>A vivid lesson on how not to be the pygmalion of your craft.</p><p><a href="https://www.youtube.com/watch?v=pGntmcy_HX8">https://www.youtube.com/watch?v=pGntmcy_HX8</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3b453c352905" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI Threat by Nouriel Roubini and Stanford AI Index Report 2023]]></title>
            <link>https://medium.com/@minhaaj/ai-threat-by-nouriel-roubini-and-stanford-ai-index-report-2023-9794b11c6c2a?source=rss-95f3a1e37177------2</link>
            <guid isPermaLink="false">https://medium.com/p/9794b11c6c2a</guid>
            <category><![CDATA[stanford]]></category>
            <category><![CDATA[china]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[economics]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Minhaaj Rehman]]></dc:creator>
            <pubDate>Tue, 11 Apr 2023 20:35:01 GMT</pubDate>
            <atom:updated>2023-04-11T20:35:01.871Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/179/1*JN7zJgj5WHdXfyqWs7BeHA.jpeg" /></figure><p>As AI investment goes down and peak of inflated expectations is going down the spiral of Trough of disillusionment, businesses are abandoning the unfulfilled promises of AI and are focusing on fundamentals and even more, the looming geopolitical mega threats.</p><p>Last week a very questionable but exhaustive AI Index report 2023 was published by <a href="https://www.linkedin.com/company/stanfordhai/">Stanford Institute for Human-Centered Artificial Intelligence (HAI)</a>. I would rather follow <a href="https://www.linkedin.com/in/ACoAAA7Fve8B363Azz8dcfBsk0Nf8VDYmxsm_Y0">Nathan Benaich</a>’s State of the AI Report and <a href="https://www.linkedin.com/company/australian-strategic-policy-institute/">Australian Strategic Policy Institute</a>’s Critical Technology report for a relatively better and less politically motivated picture.</p><p>Nonetheless here are some emerging themes in the report:</p><p>1. Industry races ahead of academia. (Academia has actually 0 contributions now as noted by better reports as opposed to what is suggested in the report)<br>2. The number of incidents concerning the misuse of AI is rapidly rising.<br>3. Performance saturation on traditional benchmarks. (Actually, there is very little to be gained on the algorithmic side, for the most part.)<br>4. According to Luccioni et al., 2022, BLOOM’s training run emitted 25 times more carbon than a single air traveler on a one-way trip from New York to San Francisco.<br>5. Global AI private investment was $91.9 billion in 2022, which represented a 26.7% decrease since 2021.<br>6. Chinese citizens are among those who feel the most positively about AI products and services. The Chinese government has the highest satisfaction rating in the world according to the IPSOS report. It is an undoubted AI leader as noted in literally every other report, that is unbiased and free of Western propaganda.</p><p>However, I personally believe in a more broader and interdependent analysis of the situation that includes externalities and macroeconomic variables. Without such dependencies, any report is too suggestive and leaves erudite analysts asking for more, in my humble opinion. This is exactly why Dr. <a href="https://www.linkedin.com/in/ACoAAAFJOcsBx6eYNfho_iTS65DnyTMr2XMVj-A">Nouriel Roubini</a>’s book Megathreats despite its dystopian outlook, is such a brilliant exposé of global economic crises and their spillover effects. He is known as Dr. Doom for his predictions of the 2008 financial crisis and his attention to detail.</p><p>In chapter 8, AI Threat he seems to encourage the idea that things that can be automated must be automated but the idea must be ensconced in pragmatism and not fiction. He summarizes the chapter with these words:</p><p>“Earth may be lucky to reach the intelligence explosion of the singularity. Will a deadly pandemic finish us before the transition to machines is complete? Will climate change destroy the planet before rational machines come to the rescue? Will we suffocate under a mountain of debt? Or will the United States and China destroy the world in a military conflict as competition to control the industries of the future becomes extreme? Indeed who controls AI may become the dominant world super-power. This geopolitical rivalry forms the megathreat we turn to next.”</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9794b11c6c2a" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>