weekend ai reads for 2024-08-09

📰 ABOVE THE FOLD: BEING HUMAN

In the Age of A.I., What Makes People Unique? / The New Yorker (10 minute read)

Vallor’s worry isn’t that artificially intelligent computers will rise up and dominate humanity, but that, faced with computers that can pretend to have human virtues, we’ll lose track of what those virtues really are. Comforted by computers that tell us that they love us, we’ll forget what love is. Wowed by systems that seem to be creative, we’ll lose respect for actual human creativity—a struggle for self-expression that can involve a “painful” reimagining of the self.

  • via adam m’s reference to The AI Mirror by Shannon Vallor (Bookshop link)

Making AI Assistants People Will Be Excited About — A human-centered approach to the next generation of AI. / Ideo (11 minute read)

Quantitative Destruction / Off Kilter (9 minute read)

Put simply; such leaders tend to lack empathy for other human beings, especially the customer, measure everything at its most rational level, revel in tiny statistical differences, elevate efficiency above all, seek to boil complex, open reality down into closed, simplistic, quantitative models, and ignore very hard or impossible-to-measure qualitative information, also known as qualia, completely.

Replacing my Right Hand with AI — Breaking my hand forced me to write all my code with AI for 2 months. I’m never going back. / Erik Schluntz (11 minute read)

Human engineers won’t go away. We’ll still be needed to drive high-level prioritization, understand the overall architecture and scope of the problem, and review the AI’s work, especially as systems get bigger. But we’ll spend much more of our time thinking about what to build, and much less on the repetitive “how” of building it.

Anthropomorphism might cause users to place more trust in the output of a model when it “hallucinates” incorrect information, OpenAI says. Over time, it might even affect users’ relationships with other people. “Users might form social relationships with the AI, reducing their need for human interaction—potentially benefiting lonely individuals but possibly affecting healthy relationships,” the document says.

 

📻 QUOTES OF THE WEEK

But if you do something well enough, and you stick to your principles, and you stick to quality and integrity, you could be on the dark side of the moon, and people will still venture for that, because people, I think, yearn for quality and they yearn for excellence.

Be Bold, action has genius.

Begin it, begin it now!

Charlie Trotter (source)

 

🏗️ FOUNDATIONS & CULTURE

Out of slightly over 200 job applications received for Pillser in the last 3 days, more than 90% have been marked as written by ChatGPT.

On one hand, the AI applications are so obvious that this acts as a sort of natural filter. On the other hand, for organizations that are not equipped to easily identify such applications, this means that good candidates will get lost in a sea of low effort applications.

Rocco Casagrande entered the White House grounds holding a black box slightly bigger than a Rubik’s Cube. Within it were a dozen test tubes with the ingredients that — if assembled correctly — had the potential to cause the next pandemic. An AI chatbot had given him the deadly recipe.

How Scotiabank Built an Ethical, Engaged AI Culture — This Canadian bank has become known for developing AI tools that are both innovative and responsible. / MIT Sloan Management Review (6 minute read)

However, organizations need to recognize that having a solid technical foundation is essential in the 21st century. This starts with having a competent CTO and extends to capable internal engineering teams. Additionally, CTOs should be on an equal footing with the CEO, rather than just reporting directly to them.

While buying AI solutions can be a quick fix and an easy start to your AI journey, it often leads to a lack of deep integration and understanding of AI within the company.

Essentially, your skull is serving as the robot’s base, and your tooth and the drill are in the same reference frame. Purely mechanical coupling means there’s no vision system or encoders or software required: it’s a direct physical connection so that motion compensation is instantaneous.

Take heed before using artificial intelligence, new ABA ethics opinion says / American Bar Association Journal (4 minute read)

So even if AI saves 20 hours by doing 15 minutes of document searching, for example, the firm should only bill for the 15 minutes of actual time worked.

For example, if a lawyer relies on a GAI tool to review and summarize numerous, lengthy contracts, the lawyer would not necessarily have to manually review the entire set of documents to verify the results if the lawyer had previously tested the accuracy of the tool on a smaller subset of documents by manually reviewing those documents, comparing then to the summaries produced by the tool, and finding the summaries accurate.

Generative AI in retail: LLM to ROI — Retailers continue to experiment with generative AI and are seeing that the technology holds great promise for reigniting growth—if they move quickly to seize the opportunity. / McKinsey & Company (17 minute read)

 

🎓 EDUCATION

There’s a Tool to Catch Students Cheating With ChatGPT. OpenAI Hasn’t Released It. — Technology that can detect text written by artificial intelligence with 99.9% certainty has been debated internally for two years / Wall Street Journal (9 minute read)

What do people really ask chatbots? It’s a lot of sex and homework — AI chatbots are taking the world by storm. We analyzed thousands of conversations to see what people are really asking them and what topics are most discussed. / Washington Post (11 minute read)

The #1 AI Expert on Earth: Dr. Andrew Ng / Ed on the Edge (33 minute audio)

Silicon Valley parents are sending their kids to AI summer camps / San Francisco Standard (8 minute read)

Some Bay Area parents are so eager to get their kids in on AI’s ground floor that they try to sneak toddlers into advanced courses. “Sometimes they’ll bring a 4-year-old, and I’m like, you’re not supposed to be here,” Du said.

 

📊 DATA & TECHNOLOGY

Q2 2024 Amazon.com Inc Earnings Call / Yahoo Finance (43 minute read)

With Q’s code transformation capabilities, Amazon has migrated over 30,000 Java JDK applications in a few months, saving the company $260 million and 4,500 developer years compared to what it would have otherwise caused. That’s the game changer.

  • if that’s even only half-true, it’s a massive understatement

‘You are a helpful mail assistant,’ and other Apple Intelligence instructions — Some pre-prompt instructions for Apple’s upcoming AI features are stored on your Mac, saying ‘do not hallucinate’ and to avoid ‘negative’ themes. / The Verge (3 minute read)

  • even Apple’s anti-hallucination strategy is to ask the LLM nicely

ScrapetoAI by Simplescraper — Extract data from websites for custom GPTs and LLMs

Receive Markdown, JSON or a CSV file in seconds that you can upload directly to OpenAI, Claude and more

  • free in the browser, and free tier for automated scraping available

FLUX: This new AI image generator is eerily good at creating human hands — FLUX.1 is the open-weights heir apparent to Stable Diffusion, turning text into images. / Ars Technica (12 minute read)

 

🎉 FUN and/or PRACTICAL THINGS

alphaXiv — Open research discussion directly on top of arXiv

ColorBliss Create printable custom coloring sheets in seconds.

Ragna Desktop — Your Private AI Chatbot for Desktop

RAGNA Desktop is a private AI multitool that runs locally on your desktop PC or laptop without an internet connection. The app is designed to help you automate repetitive tasks, increase efficiency, and free up capacity for the really important things.

  • so far, less customizable and extensible than Ollama, but also more user-friendly; your mileage may vary

  • includes AI-powered text editor

Arte — Your free AI tax research assistant

Toolstash — Track, lend & borrow tools

  • the “Get DIY help right away based on the tools you have” feature has real potential

  • iOS-only

  • FYI: they gather a lot of information when you use their app

  • until the Olympics, we didn’t know we needed an AI-powered suspension system

With this combination of sensors, Flight Attendant builds a picture of the terrain and the rider’s pedalling input. Based on that picture, it automatically adjusts the suspension to the ideal setting. Put simply, it’s designed to firm up the suspension to improve pedal efficiency on the climbs and along smoother terrain, while allowing the suspension to open up for the descents and on rougher trails. And all without your hands ever having to leave the grips.

 

🧿 AI-ADJACENT

Raw intuitions about startups / Gordon Brander, Squishy, Substack (sorry) (13 minute read)

The sweet spot for startups is seeing 24 months into the future. This is a stone’s throw into the adjacent possible. It’s far-out enough to be asymmetric, familiar enough to be pitchable, and time enough to build.

…But you have to build fast! The frontier of the adjacent possible is always expanding.

source: “Possible futures, preferable futures”, Bezold Hancock (1994)