Weekend Briefing: AI's Enterprise Reality Check & the Privacy Cost of Progress
Your weekly digest of AI breakthroughs, flying taxi failures, and tech's unintended consequences
Dear Friend—
This week was a capital-P week in the world of PESTLE – most of the other letters in the alphabet seemingly took a backseat. This being said, there is still enough going on in our future world – let’s dig in!
Have a beautiful weekend!
P ツ
The Thin Wisps of Tomorrow
VC firm Menlo Ventures released their 2024: The State of Generative AI in the Enterprise report. There is lots of good information in there—the slide which stood out to me is this:
The vast amount of use cases for AI are still confined to the most obvious ones (not surprisingly—but still): Code generation (GitHub Copilot and Co), Support Chatbots (I doubt many of those will be fully fledged GenAI-powered as it is still too risky to let the wider public interact with an LLM unfiltered), anything to do with data search and retrieval and the generation of text. The truly juicy use cases, which have potential to actually save large amounts of money and/or transform the business aren’t there yet.
Headlines from the Future
🍲 Signs of the times - Voting opens for Oxford Word of the Year 2024, with one of the words being “slop” (AI-generated spam).
🚁 Another case of the “if you had paused for a moment, you would have realized how stupid this all is” (the same is true for the Metaverse): Europe’s flying taxi dreams falter as cash runs short.
👨🏼🔬 Fascinating research paper from MIT on the huge impact of AI on scientific discovery (+44% discovery of new materials, +39% patent filings, and +17% downstream product innovation): Artificial Intelligence, Scientific Discovery, and Product Innovation.
🔄 Counterpoint to the “AGI is just around the corner”-narrative: What if AI doesn’t just keep getting better forever?
🛒 Never underestimate the (un)intended consequences: Banning free plastic bags for groceries resulted in customer purchasing more plastic bags, study finds.
🤫 Here is Elon’s (not so super secret) secret five step plan on how to GSYHIDO (Getting your S#!% Done): Elon Musk’s Engineering Principles.
🚗 Who would have thought that the moment we add more connected tech to our cars, those cars would become a driving privacy nightmare?! Modern cars are surveillance devices on wheels with major privacy risks.
🥨 Remember when we talked about the German grocery chain Lidl beating Google, Amazon, and Microsoft at their own game of running data centers and making them available to other companies? Well, now Google joins forces with Lidl: Schwarz Group partners with Google on EU sovereign cloud.
What We Are Reading
🌊 How Amazon And IKEA Are Pushing The Ocean Shipping Industry To Meet Net-zero Emission Goals In an unexpected alliance, retail rivals Amazon and IKEA are joining forces to revolutionize ocean shipping with clean e-fuels, proving that competitive giants can unite for planetary change. @Jane
🕵️♂️ The Business-School Scandal That Just Keeps Getting Bigger Business school psychology research keeps getting tagged as fraudulent, with decorated professors being exposed for fabricating data. The craziest part of all is that one of the fraudulent papers is about the human tendency to misreport facts and figures for personal gain. @Mafe
🧬 Levers For Biological Progress A long, fascinating read on what the AI boom can (and can’t) do to accelerate the progress of experimentation and innovation in biology. @Jeffrey
🌱 This App Set Out To Fight Pesticides. After VCs Stepped In, Now It Helps Sell Them Plantix, originally designed to help farmers diagnose crop issues and reduce pesticide use, now focuses on enabling farmers to buy pesticides. The focus shifted once VC funding became part of the conversation. @Pedro
📝 In The Context Of Long Context One of the biggest drivers of usefulness in current generation LLMs isn’t necessarily the size of their training set, nor the specific algorithms, but the vastly expanding context windows (i.e., how much data we can feed an AI). That’s pretty much the sole reason NotebookLM from Google is so great. @Pascal
The Fun Stuff
🎮 Remember the good old days of nerdy nights of gaming in your frat house? They are coming back (for some): We built our house for LAN parties.
🦇 Thought echolocation is only for bats? Think again and train yourself: Teach Yourself to Echolocate.