Liked Stuff we figured out about AI in 2023 (simonwillison.net)

2023 was the breakthrough year for Large Language Models (LLMs). I think it’s OK to call these AI—they’re the latest and (currently) most interesting development in the academic field of Artificial Intelligence that dates back to the 1950s.

Here’s my attempt to round up the highlights in one place!

Source: Stuff we figured out about AI in 2023 by Simon Willison

Replied to Whose ethics? Whose AI? (helenbeetham.substack.com)
Helen Beetham follows up here keynote to the Association for Learning Technologies (ALT) winter summit on AI and ethics, unpacking a few questions that were raised. The two points that stood out to me were that rather than teaching ‘prompt engineering‘, we need to update search skills in a part-AI world:

What I think we probably should do, working with our colleagues in libraries and study skills centres, is to update our support for search skills. Help students to understand what the algorithms are hiding as well as what they are revealing, how to search when you know what you are looking for as well as when you don’t, the business models as well as the algorithms of search, and how search online is being systematically degraded both by commercial interests and by these new synthetic capabilities. We may conclude that students are better off learning how to use the walled gardens of content that academic libraries and subscriptions provide – which ties in with my arguments about building our own ecosystems. That will also equip them for search ‘in the wild’.

Source: Whose Ethics? Whose AI? by Helen Beetham

And that rather than inviting cynicism regarding the use of various tools to write essays, opportunities need to be found to talk to students about what agency they have to shape their own futures.

ᔥ “Stephen Downes” in Downes.ca ~ Stephen’s Web ~ Whose ethics? Whose AI? ()

Replied to The digital wall by David TrussDavid Truss (daily-ink.davidtruss.com)

What is it about the internet that gives people permission to be awful and mean to others? I follow an astrophysicist on social media. She’s brilliant, and makes great content. She also posted a rant about all the misogynistic comments she gets from men commenting on her rather than her content. I’m not sharing any more details because it looks like she took the video down.

Source: The Digital Wall by David Truss

David, I have long wondered about the problem of on and offline. As an educator, are there any strategies or approaches that you have put in place to encourage empathy online, as well as an understanding of the impact of such practices? I recently did a short course on cyber security and awareness, my feeling is that such comments risk forming an informal character reference in a world beyond forgetting.

I think the future of hacking and cyber attacks is the linking of different datasets that we openly share online through data brokers to provide an insight and awareness of individuals that will open up new possibilities.

Source: Cyber Security & Awareness – Primary Years (CSER MOOC) by Aaron Davis

Ironically, looking back through my blog I actually came across a previous post and comment on your blog relating to the difference between our online and offline persona.

I can see that we are not our online personas. They are different than us. Yet they can say a lot about… but they don’t always say what we think they say.

Source: Our Online Persona by David Truss

Replied to epilepticrabbit (@epilepticrabbit@social.coop) (social.coop)

Idk I guess I just need to go back to MailPoet. It’s going to suck for @dajb and whomever else gets auto-unsubscribed randomly, but I spent almost 3 hours looking at alternatives and my newsletter on my site seems like the best thing to do.

Laura, I received the first Mailpoet newsletter if that means anything. Really appreciated your point about tech principles:

I have inconvenienced myself with my tech principles for decades at this point. It takes days for my local bookshops to send me books because I try very hard to not use Amazon. I am not a part of several family and friend What’sApp groups because I loathe anything Facebook is or touches. I have, for decades, tried to live up to my values, fought for the open web and made technology choices that align with who I am and what I believe in.

Source: %5BFBT%5D%20on%20MailPoet%20and%20Moving by Laura Hilliger

Liked Why Oblivion.University? by mbransonsmbransons (michaelbransonsmith.net)

This post is a component of my coursework at the CUNY Graduate Center in the M.S. in Data Analysis and Design, for a special topics course about Large Language Models taught by Professor Michelle A. McSweeney. I extend my gratitude to her and all my fellow students for embarking on this journey together, seeking to […]

Oblivion.University enables users to submit text-based work and questions for Dr. Oblivion’s AI-generated feedback. The feedback, presented in the form of an AI-generated video embed (currently audio only), undergoes a multi-step process. This process involves a fine-tuned LLM response, audio transformation using a voice clone, and video rendering with a lip-syncing generative model (the final step is currently only achievable offline).

Source: Why Oblivion.University? by Michael Branson Smith

Replied to https://johnjohnston.info/blog/books-2023/ by john john (johnjohnston.info)

Montage of coverts of 7 books in a strip: Demon Copperhead Hungry Ghosts The Apparition Phase The Secret History The Seven Moons Tomorrow and Tomorrow and Tomorrow You Have a Friend in 10A
According to my Books page, I read 43 books last year, three less than the year before year. A list of 5 star books (weight by my enjoyment, pretty vague) using the lovely display posts plugin. Read Demon CopperheadRead: Hungry GhostsRead: You Have a Friend in 10ARead: The Apparition PhaseRead: The Secret HistoryRead: The Seven Moons of Maali Almeida
 
 

Happy new year John. Is Display Posts a combination of ShortCode and code snippets added to your theme or just ShortCode? Also, is you books page created using Display Posts too?
https://en.wikipedia.org/wiki/Videodrome
I had never seen Videodrome. It felt like a film that leaves many questions. It made me want to dig back into Deleuze and the body without organs.

The concept of the body without organs was mainly defined by Deleuze and Guattari in the two volumes of their work Capitalism and Schizophrenia, Anti-Oedipus and A Thousand Plateaus.[11] In both books, the abstract body is defined as a self-regulating process—created by the relation between an abstract machine and a machinic assemblage “Assemblage (philosophy)”)—that maintains itself through processes of homeostasis and simultaneously limits the possible activities of its constituent parts, or organs.[12] The body without organs is the sum total intensive and affective activity of the full potential for the body and its constituent parts.[13]

Source: Body Without Organs by Wikipedia

ᔥ “Jim Groom” in AI106: Long Live the New Flesh | bavatuesdays ()

Watched 2018 documentary film directed by Peter Jackson by Contributors to Wikimedia projects from Wikimedia Foundation, Inc.

They Shall Not Grow Old is a 2018 documentary film directed and produced by Peter Jackson. The film was created using original footage of the First World War from the Imperial War Museum‘s archives, most previously unseen, all over 100 years old by the time of release. Audio is from BBC and Imperial War Museum (IWM) interviews of British servicemen who fought in the conflict. Most of the footage has been colourised and transformed with modern production techniques, with the addition of sound effects and voice acting to be more evocative and feel closer to the soldiers’ actual experiences.

Source: They Shall Not Grow Old by Wikipedia

They Shall Not Grow Old is documentary that ties together voices from the war with archival film and images from World War One. It was co-commissioned by 14–18 NOW and Imperial War Museums in association with the BBC, with a copy sent to every school in the United Kingdom.

The film was directed by Peter Jackson. He used technology to bring new life to original footage from World War One, fixing up the inconsistencies with the rates per minute and adding colour. There are also moments when actors have been brought in to add voice to the silent film with the scripts produced by professional lip readers who watched the archival material.

As a narrative, it begins with the build-up to war and the excitement about enlisting, even if you were not officially old enough. It then recounts the training for soldiers and the journey to Europe and the trenches. We are given an insight into life in the trenches, including the maze like structure, how you rested where you were, the food eaten, such as bread, bacon, biscuits and bully beef, where people went to the toilet, and how soldiers dealt with infestations of lice and rats. This is contrasted with constant shelling, gas attacks, and the chaos of going over the top to take an enemy line. It then ends with armistice and the neglected and misunderstood life of the returned serviceman.

Overall, the film ties together different facets of war into an odd narrative about the western front that seemingly existed for so many, but for no-one in particular. I think that it is telling how much material was reviewed for the project.

The crew reviewed 600 hours of interviews from 200 veterans and 100 hours of original film footage to make the film.

Source: They Shall Not Grow Old by Wikipedia

Although there has been a lot of praise for the film.

Jackson has done something quite remarkable: using 21st-century technology to put the humanity back into old movie stock.

Source: They Shall Not Grow Old review – an utterly breathtaking journey into the trenches by Mark Kermode

There are also some who think that we need to be mindful of the choices made and the act of history making.

But the colourisation combined with the selective source base, the implicit narrative making and the critical response that suggests that this is somehow more “authentic” history, is problematic. Some reviewers seem unable to distinguish fiction from reality: “No Lord of the Rings battle could match the sheer hellishness of what the filmmaker recreates here,” writes one.

What does this process of modernisation and the addition of colour and sound, which Jackson advocates for wider usage across historical archives, do for our understanding of the past? On Armistice Day, we should encourage people to watch this film – not just for its World War I history, but as a good opportunity to think about history making.

Source: They Shall Not Grow Old: World War I film a masterpiece of skill and artistry – just don’t call it a documentary by Alice Kelly

Replied to AI + Google Sheets: How To Use Them Together by Ben Ben (benlcollins.com)

See how AI tools work with Google Sheets to boost your productivity. Covers ChatGPT, Google Bard, and AI add-ons.

I really enjoyed how you broke down the different uses for AI in this post Ben. Personally, I have found myself using Co-Pilot to come up with formulas as this is what I currently have access to. For me, I often have an idea of what is possible, but do not always have the time or mental space to dig into the formulas to find the right solution. I have found it useful then to just ask Bing. I am now finding myself teaching colleagues how to use prompts to not only find a solution, but have it explained for them.
Listened Wisenheimer album by Custard by Contributors to Wikimedia projects from Wikimedia Foundation, Inc.

Wisenheimer is the third studio album by the Australian band Custard. It was released November 6th 1995 and peaked at number 55 in September 1996. The album contains the song “Apartment” which reached #7 in the third Hottest 100.

I wonder how many people came to Wisenheimer, Custard’s third album, after hearing the opening single, Apartment, and were somewhat disappointed? This is the feeling that I get from Lachlan J’s review:

So, my verdict is that Custard’s Wisenheimer is a pretty good album. It’s not great and it doesn’t really capture my imagination in any significant way, but it is quite a bit of fun to let play while I’m driving about the town or doing the housework. It has some rather good tracks on it and it has quite a few rather average tracks, but it doesn’t really have any bad tracks, which is a nice upside. The song-writing is decent and the musicians are competent, but it’s really nothing to write home about, and the scope of music is comfortably broad, but nothing particularly challenging or intriguing. Really this album is just a nicely comfortable piece of work. It doesn’t break any boundaries, but it’s good enough.

Source: Wisenheimer – Custard by Lachlan J

This is something that the band’s manager, Dave Brown, touches on in Andrew Stafford’s Pig City, arguing that Apartment was released too early:

Custard had met keyboard player and producer Eric Feldman while touring in support of another of McCormack’s heroes, former Pixie Frank Black. Feldman had a long list of credits and contacts, and Frank Black himself had been impressed enough by Custard to loan McCormack three of his guitars for the recording of what was to become Wisenheimer. If the album lacked its predecessor’s rambling charm, it also contained some brilliant material (the woozy, beautiful art-rock of Columbus is perhaps Custard’s greatest moment).

The obvious standout, Apartment, was the first single. It was a disappointing choice for Dave Brown, who reasoned that by leading with their best punch, excellent follow-up singles such as Lucky Star and Sunset Strip were rendered anti-climactic after the album’s release in late 1995. Dave Brown:

It’s always my bitch that they released Apartment at the wrong time, and that was the difference between Wisenheimer being a successful album versus a really successful album. It was the first single and it was too good for that, without a doubt in the world. It should have been released second or third; I think that gets proven every time.

Source: Pig City by Andrew Stafford

Comparing the album with Wahooti Fandango, I kept on thinking that having one producer for the whole album, Eric Drew Feldman, made it more consistent, but I feel that is possibly in the ear of the beholder. Maybe, Wisenheimer is less contrasting than Wahooti Fandango, but each track still jumps around between genres, whether it be the angular rock guitar one minute with GooFinder, to leaning back into the country origins with Leisuremaster. There are also strange interludes and extras, such as the saxophone led jam of Cut Lunch or the the excerpt about gold at the end of I Love Television that reminded me of Jim Carey’s monologues on The Weeknd’s Dawn FM.

With the length of tracks, I feel that you never really get to settle as a listener. Even the slower tracks fly on by.  Or maybe like a box of Roses chocolates, this is an album for those who just like eating chocolates, no matter the flavour, but would possibly frustrate those who just like this flavour or that. I wonder this maybe what Damian Cowell was touching upon when he spoke about Custard and anchovies. All in all, it was one of those albums that really benefited from multiple plays.

On a side note, the one thing that I am left intrigued by is how they presented this tapestry of sounds live? The sound often contrasts between a wall of sound and more subtle sounds. When I saw McCormick live playing acoustically, it felt like the tracks were chosen because they fitted the bill, with the only track that felt like it did not fit was Girls, but nobody cared. However, thinking about it now, I wonder if McCormick / Custard could in fact play a number of different sets that would cater for different audiences? I have searched YouTube in the vain hope of finding an old concert, but all I can find is them performing Apartment.

Bookmarked How do I remove the same part of a file name for many files in Windows 7? (Super User)

get-childitem *.mp3
This lists all files whose names end with .mp3. They are then piped to the next command with the | operator.

foreach { rename-item $_ $_.Name.Replace(“Radiohead -“, “”) }
This replaces all instances of Radiohead – with nothing, denoted by “”, effectively wiping the word from all the files in the directory.

You could also modify get-childitem *.mp3 to get-childitem – that would rename all the files in the directory, not just files whose names end with .mp3.

I was cleaning up my phone and find some files that I wanted to update in bulk. I knew that it was doable, but had never properly delved into this world (or Googled it). I found this guide to using PowerShell to do it.

As I go further into the world of SQL and scripting, I can see how a lot of my processes are going to change.

Replied to How Britain’s taste for tea may have been a life saver by Veronique Greenwood (BBC)

Tea became one of the British Empire’s most prized resources in the 18th Century. But it may have also had an unintended effect on the British population – reducing mortality rates.

In a recent paper in the Review of Statistics and Economics, economist Francisca Antman of the University of Colorado, Boulder, makes a convincing case that the explosion of tea as an everyman’s drink in late 1700s England saved many lives. This would not have been because of any antioxidants or other substances inherent to the lauded leaf.

Instead, the simple practice of boiling water for tea, in an era before people understood that illness could be caused by water-borne pathogens, may have been enough to keep many from an early grave.

Source: How Britain’s taste for tea may have been a life saver
 by Veronique Greenwood

I was expecting the benefit of tea might be less drinking of beer, I guess boiling water does make a lot of sense.

Replied to How to Use Technology for Documentation in the Classroom (Zoe Porter-Children and Technology)

Technology has been integrated into the classroom over the past several years. Teachers use it to teach their students new concepts. What about documentation? Teachers also need to keep data about how the students in their class are developing. There are certain milestones administration expects children to reach throughout the school year. Keeping the data about these milestones is an important part of the work of educators. Teachers can do this in different ways. Some teachers may write informal notes about how children are doing when they see a specific behavior. Other teachers may use more formal methods, like keeping detailed observations in a file on their computer. Technology can be used to keep the data teachers collect on their students organized.

I am glad that you found my investigations into the use of technology to support documentation helpful. In case you are interested, I found Making Thinking Visible by Ron Ritchhart, Mark Church and Karin Morrison, a useful read and reference, as well as all the Project Zero resources. One of the things that I was always challenged by was the importance of ‘advancing learning’, not just capturing it.

The focus throughout is the development of understanding, rather than as some sort of by-product. Central to this is the notion of documentation. This can be split into four practices: observing, recording, interpreting and sharing. What is important about documentation is that it, “must serve to advance learning, not merely capture it. As such, documentation includes not only what is collected but also the discussions and reflections on those artifacts.” (Page 38)

Source: An Introduction to Making Thinking Visible – Read Write Respond by Aaron Davis

Another book that I have dived in and out of on the topic has been The Hundred Languages of Children: The Reggio Emilia Experience in Transformation, a book that brings together a number of voices on various topics, including Part III, which is on documentation.

Bookmarked How Big is YouTube? (ethanzuckerman.com)
Ethan Zuckerman shares some reflections on a recent article focused on measuring how big YouTube is.

YouTube is one of the largest, most important communication platforms in the world, but while there is a great deal of research about the site, many of its fundamental characteristics remain unknown. To better understand YouTube as a whole, we created a random sample of videos using a new method. Through a description of the sample’s metadata, we provide answers to many essential questions about, for example, the distribution of views, comments, likes, subscribers, and categories. Our method also allows us to estimate the total number of publicly visible videos on YouTube and its growth over time. To learn more about video content, we hand-coded a subsample to answer questions like how many are primarily music, video games, or still images. Finally, we processed the videos’ audio using language detection software to determine the distribution of spoken languages. In providing basic information about YouTube as a whole, we not only learn more about an influential platform, but also provide baseline context against which samples in more focused studies can be compared.

Source: Dialing for Videos: A Random Sample of YouTube by Ryan McGrady, Kevin Zheng, Rebecca Curran, Jason Baumgartner and Ethan Zuckerman

The information is captured in the site that created TubeStats and is updated regularly.

Separately, Ryan McGrady has summarised some key takeaways on the Initiative for Digital Public Infrastructure site:

  • There are about 10 13 billion publicly visible videos
  • YouTube is mostly not in English
  • Our current best estimate is that 32% of videos where we can detect the language are in English, with 10.5% in Hindi, 8% in Spanish, slightly fewer in Portuguese, and just over 6% in Arabic.
  • Most of YouTube doesn’t get many views

  • Not everyone is participating in the “creator economy”

  • There are an awful lot of video games

Source: 5 Main Takeaways from Randomly Sampling YouTube by Ryan McGrady

What is just as interesting as the statistics, but how they managed to capture the data through ‘drunk dialing’:

That bit after “watch?v=” is an 11 digit string. The first ten digits can be a-z,A-Z,0-9 and _-. The last digit is special, and can only be one of 16 values. Turns out there are 2^64 possible YouTube addresses, an enormous number: 18.4 quintillion. There are lots of YouTube videos, but not that many. Let’s guess for a moment that there are 1 billion YouTube videos – if you picked URLs at random, you’d only get a valid address roughly once every 18.4 billion tries.

We refer to this method as “drunk dialing”, as it’s basically as sophisticated as taking swigs from a bottle of bourbon and mashing digits on a telephone, hoping to find a human being to speak to. Jason found a couple of cheats that makes the method roughly 32,000 times as efficient, meaning our “phone call” connects lots more often. Kevin Zheng wrote a whole bunch of scripts to do the dialing, and over the course of several months, we collected more than 10,000 truly random YouTube videos.

Source: How Big is YouTube? by Ethan Zuckerman

After reading Jim Groom’s post about an AI Dr Oblivion, I am left wondering about what the numbers really mean.

Replied to What is the purpose of educational technology? (blog.edtechie.net)

I don’t mean that title as a rhetorical, smartass, question, but rather a more fundamental one. It’s probably not one we ask ourselves very often, we tend to be caught up in the application of a particular technology, or trying to solve a specific problem. But at the more abstract level, what do you think educational technology is for? When we adopt it, what is the purpose we are intending it to fulfil? I expect the answer will vary depending on technology or context, and not be limited to one function overall. But of you had to answer the question “what is the main purpose of educational technology?” at a cocktail party, what would you answer (apart from asking yourself how did you get invited to a party where this is the conversation).

Here are some potential responses I think:

  • Improve learning performance
  • Making learning more accessible/flexible
  • Financial benefits
  • Student experience and choice
  • Improved pedagogy
  • Reflection on practice
  • Administration and monitoring

Source: What is the purpose of educational technology? by Martin Weller

Martin, this has me thinking about a post I wrote too many years ago about my ‘vision for eLearning‘:

Is Transformative: More than just redefined, learning is purposeful and involves wider implications.

Is More Doable: Makes things like critical thinking and collaboration more possible.

Enables Student Voice: Technology provides a voice for students to take ownership over their work and ideas.

Involves Modelling Digital Citizenship: More than a sole lesson, eLearning should be about foster competencies throughout the curriculum.

Source: Vision for eLearning by Aaron Davis

The thing that struck me then was that vision is a collective enterprise and so often is contested, no matter how much work is done. For example, this week, I got caught talking with a colleague who argued that there are three facets to the technology project that we are a part of: finance, student/community and pedagogy. The problem that we face is that there is nobody who is properly invested in all the areas, therefore any decisions made are always made based on the priorities of the group in question. When I started too many years ago now, my focus was all about the students, as outlined in my vision, and although this is still the case and will always be something of a north star, my day-to-day focus these days is on administration and finance. Sadly, I have come to learn the reality that when it comes to technology at scale (financial benefit, you might say), the focus becomes the quality of data you are working with and improving the steps to producing such data more efficient. Many prefer to call this ‘magic‘, but to me it is the foundation that allows the house to be built. Invisible to most, until a massive crack appears in the wall and you need to go digging.

I was also left thinking about Ewan McIntosh’s post about the various purposes alongside Ewan McIntosh’s discussion of a school’s ‘value proposition‘. He posited that beyond two values, teams get lost:

A value proposition, even if you are a state school, is a vital value to hone down, not just so that kids aren’t ripped out of your school but so that everyone, including the leaders, can be held to account when kinks in the system appear. If you state that excellence in education is your value proposition, then you’d better get that nailed, all the time, every time, or perceptions will change and take a long time to bring back.

And defining a value proposition is easy – you can really only choose one top value you pursue, and a close-place second one. Beyond two core value propositions, your team will be lost and not know what they are chasing

Source: Working out a school’s competitive position even when it’s not competing #28daysofwriting by Ewan McIntosh

All in all, your post has me intrigued to think about what has changed and what remains the same regarding education technology.

Bookmarked The Linkfest (buttondown.email)

The opposite of doomscrolling: Every week (roughly) I send you a collection of the best Internet reading I’ve found — links to culture, technology, art and science that fascinated me.

Linkfest is Clive Thompson’s newsletter of interesting tidbits from the World Wide Web. I can only imagine the time and effort that goes into something that seems so fleeting. I also like his recognition that ‘it takes a village’:

I read hundreds of blogs and news sites every week to find this stuff. Shoutout to one I relied on particularly this week — Charles Arthur’s The Overspill. Go check it out!

Source: Linkfest #15: Altruistic pigs, the “Truetown Discharge”, and the CIA’s guide to wrecking meetings by Clive Thompson

Stephen, thank you for the reminder of what I have missed this year after taking a step back from things. I felt that so much of my online life had become stale, repeatable and templated, I wondered if it ‘sparked joy‘ anymore. I had wondered if I was doing things out of habit, rather than with purpose. Sporadically, diving back into my feed, I added a few posts to my site, one on the AI bubble and the other on educational communities, only to discover your responses:

Let’s allow that AI is a bubble (this saves us an exercise in semantics). Is it true that all bubbles pop? Yes, we can name many bubbles that have popped, but let’s consider some other technologies that experienced rapid growth, so much so that any ‘pop’ of the bubble was merely a rounding error. Like, say, the telephone. Cars. Aircraft. Microwave ovens. Computers. I could go on, but you get the idea. We tend to forget the bubbles that didn’t pop, because they became fixtures of everyday life.

Source: Pluralistic: What kind of bubble is AI? by Stephen Downes

And …

Twitter was only ever a subset of the larger educational community, and it always felt to me a bit like a high school clique. It was pretty easy to find yourself on the outside or the subject of Twitter disapprobation; I experienced it a few times (but to nowhere near the extent of some others). If Twitter was the best we could hope for, we weren’t hoping for much.

Source: Communities and Conversations of the Past by Stephen Downes

They were like seeds sprouting in the garden. As always, you added perspective that spurred me to think more, a reminder of the interactions I had missed diving into the world of books.

Liked Pluralistic: What kind of bubble is AI? (19 Dec 2023) by Cory DoctorowCory Doctorow (pluralistic.net)

Once the bubble pops (all bubbles pop), AI applications will have to rise or fall on their actual merits, not their promise. The odds are stacked against the long-term survival of high-value, risk-intolerant AI applications.

The problem for AI is that while there are a lot of risk-tolerant applications, they’re almost all low-value; while nearly all the high-value applications are risk-intolerant. Once AI has to be profitable – once investors withdraw their subsidies from money-losing ventures – the risk-tolerant applications need to be sufficient to run those tremendously expensive servers in those brutally expensive data-centers tended by exceptionally expensive technical workers.

If they aren’t, then the business case for running those servers goes away, and so do the servers – and so do all those risk-tolerant, low-value applications. It doesn’t matter if helping blind people make sense of their surroundings is socially beneficial. It doesn’t matter if teenaged gamers love their epic character art. It doesn’t even matter how horny scammers are for generating AI nonsense SEO websites

Source: What kind of bubble is AI? by Cory Doctorow

Liked Google’s True Moonshot (Stratechery by Ben Thompson)

Google could do more than just win the chatbot war: it is the one company that could make a universal assistant. The question is if the company is willing to risk it all.

What, though, if the mission statement were the moonshot all along? What if “I’m Feeling Lucky” were not a whimsical button on a spartan home page, but the default way of interacting with all of the world’s information? What if an AI Assistant were so good, and so natural, that anyone with seamless access to it simply used it all the time, without thought?

Source: Google’s True Moonshot by @stratechery