Dynamic Datalist: Autocomplete from an API :: Aaron Gustafson
Great minds think alike! I have a very similar HTML web component on the front page of The Session called input-autosuggest.
Link tags: toc
87
Curate your own newspaper with RSS
I’m almost certainly preaching to the choir here because I bet you’re reading these very words in a feed reader, but what Molly White has written here is too good not to share:
RSS offers readers and writers a path away from unreliable, manipulative, and hostile platforms and intermediaries. In a media landscape dominated by algorithmic feeds that aim to manipulate and extract, sometimes the most radical thing you can do is choose to read what you want, when you want, without anyone watching over your shoulder.
AI doesn’t need to think. We do! - craigabbott.co.uk
A good overview of how large language models work:
The words flow together because they’ve been seen together many times. But that doesn’t mean they’re right. It just means they’re coherent.
An Ars Technica history of the Internet, part 1 - Ars Technica
Here’s a fun account of the early days of the ARPANET.
[this is aaronland] a tale of gummy snakes (and spunk)
About halfway through this talk transcript, Aaron starts dropping a barrage of truth bombs:
I understand the web, whose distinguishing characteristic is asynchronous recall on a global scale, as the technology which makes revisiting possible in a way that has genuinely never existed before the web.
What the web has made possible are the economics of keeping something, something which has not enjoyed “hockey stick growth”, around long enough for people to warm up to it. Or to survive long past the moment when people may have grown tired of it.
If your goal is to build something which is designed to flip inside of ten years, like many things in the private sector, that may not seem like a very compelling argument.
If, however, your goal is to build something to match the longevity of the cultural heritage sector, to meet the goal of fostering revisiting, or for novel ideas to outlast the reluctance of the present and to do so at a global scale, or really any scale larger than shouting distance, then I will challenge you to find a better vehicle for doing so than the internet, and the web in particular.
What I’ve learned about writing AI apps so far | Seldo.com
LLMs are good at transforming text into less text
Laurie is really onto something with this:
This is the biggest and most fundamental thing about LLMs, and a great rule of thumb for what’s going to be an effective LLM application. Is what you’re doing taking a large amount of text and asking the LLM to convert it into a smaller amount of text? Then it’s probably going to be great at it. If you’re asking it to convert into a roughly equal amount of text it will be so-so. If you’re asking it to create more text than you gave it, forget about it.
Depending how much of the hype around AI you’ve taken on board, the idea that they “take text and turn it into less text” might seem gigantic back-pedal away from previous claims of what AI can do. But taking text and turning it into less text is still an enormous field of endeavour, and a huge market. It’s still very exciting, all the more exciting because it’s got clear boundaries and isn’t hype-driven over-reaching, or dependent on LLMs overnight becoming way better than they currently are.
Fine-tuning Text Inputs
Garrett talks through some handy HTML attributes: spellcheck, autofocus, autocapitalize, autocomplete, and autocorrect:
While they feel like small details, when we set these attributes on inputs, we streamline things for visitors while also guiding the browser on when it should just get out of the way.
“Wherever you get your podcasts” is a radical statement - Anil Dash
What podcasting holds in the promise of its open format is the proof that an open web can still thrive and be relevant, that it can inspire new systems that are similarly open to take root and grow. Even the biggest companies in the world can’t displace these kinds of systems once they find their audiences.
AI Art is The New Stock Image
While some executives in Davos may get excited about its infinite possibilities this week, to a younger consumer AI Art is already ‘a bit cringe’.
A cohesive and unified identity for IndieWeb protocols · Paul Robert Lloyd
I really enjoyed hanging out with Paul at Indie Web Camp in Nuremberg last weekend. And I like the iconography he’s proposing:
This design attempts to bring together a set of icons that share the concept of a node – a line and a point – and use this to add counters to each letter shape.
Vibe Shift
Forget every article you’ve read that tries to explain large language models. Just read this post by Peter and feel it.
ActivityPub is the next big thing in social networks - The Verge
After nearly two decades of fighting for this vision of the internet, the people who believed in federation feel like they’re finally going to win. The change they imagine still requires a lot of user education — and a lot of work to make this stuff work for users. But the fundamental shift, from platforms to protocols, appears to have momentum in a way it never has before.
Stochastic Parrots Day Tickets, Fri, Mar 17, 2023 at 8:00 AM | Eventbrite
This free event is running online from 3pm to 7pm UK time this Friday. The line-up features Emily Bender, Safiya Noble, Timnit Gebru and more.
Since the publication of On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?🦜 two years ago, many of the harms the paper has warned about and more, have unfortunately occurred. From exploited workers filtering hateful content, to an engineer claiming that chatbots are sentient, the harms are only accelerating.
Join the co-authors of the paper and various guests to reflect on what has happened in the last two years, what the large language model landscape currently look like, and where we are headed vs where we should be headed.
A year of new avenues
All along, from the frothy 1990s to the percolating 2000s to the frozen 2010s to today, the web has been the sure thing. All along, it’s been growing and maturing, sprouting new capabilities. From my vantage point, that growth has seemed to accelerate in the past five years; CSS, in particular, has become incredibly flexible and expressive. Maybe even a bit overstuffed — but I’ll take it.
For people who care about creating worlds together, rather than getting rich, the web is the past and the web is the future. What luck, that this decentralized, permissionless system claimed a position at the heart of the internet, and stuck there. It’s limited, of course; frustrating; sometimes maddening. But that’s every creative medium. That’s life.
The Thorny Problem of Keeping the Internet’s Time | The New Yorker
This story of the Network Time Protocol hammers home the importance of infrastructure and its maintenance:
Technology companies worth billions rely on open-source code, including N.T.P., and the maintenance of that code is often handled by a small group of individuals toiling away without pay.
The Real Novelty of the ARPANET
Setting the scene:
The Washington Hilton stands near the top of a small rise about a mile and a half northeast of the National Mall. Its two white-painted modern facades sweep out in broad semicircles like the wings of a bird. The New York Times, reporting on the hotel’s completion in 1965, remarked that the building looks “like a sea gull perched on a hilltop nest.”
The hotel hides its most famous feature below ground. Underneath the driveway roundabout is an enormous ovoid event space known as the International Ballroom, which was for many years the largest pillar-less ballroom in DC. In 1967, the Doors played a concert there. In 1968, Jimi Hendrix also played a concert there. In 1972, a somewhat more sedate act took over the ballroom to put on the inaugural International Conference on Computing Communication, where a promising research project known as the ARPANET was demonstrated publicly for the first time.
It turns out that the most important innovation of the ARPANET isn’t obvious in hindsight:
So what I’m trying to drive home here is that there is an important distinction between statement A, “the ARPANET connected people in different locations via computers for the first time,” and statement B, “the ARPANET connected computer systems to each other for the first time.” That might seem like splitting hairs, but statement A elides some illuminating history in a way that statement B does not.
Letter in Support of Responsible Fintech Policy
A well-written evisceration of cryptobollocks signed by Bruce Scheier, Tim Bray, Molly White, Cory Doctorow, and more.
If you’re a concerned US computer scientist, technologist or developer, you’ve got till June 10th to add your signature before this is submitted to congress.
Cautionary Tales from Cryptoland
This quote from the brilliant Molly White is about web3/blockchain/cryptobollocks but it applies to evaluating technology in general (like, say, JavaScript frameworks):
I firmly believe that companies first need to identify and research the problem they are trying to solve, and then select the right technology to do it. Those technologies may not be the latest buzzword, and they may not cause venture capitalists to come crawling out of the woodwork, but choosing technologies with that approach tends to be a lot more successful in the long run — at least, assuming the primary goal is to actually solve a problem rather than attract VC money.
Notes from a gopher:// site - daverupert.com
The result of adding more constraints means that the products have a broader appeal due to their simple interface. It reminds me of a Jeremy Keith talk I heard last month about programming languages like CSS which have a simple interface pattern:
selector { property: value }. Simple enough anyone can learn. But simple doesn’t mean it’s simplistic, which gives me a lot to think about.
It’s not still the early days
If you’re interested in so-called web3, you should definitely follow Molly White.
How long can it possibly be “early days”? How long do we need to wait before someone comes up with an actual application of blockchain technologies that isn’t a transparent attempt to retroactively justify a technology that is inefficient in every sense of the word? How much pollution must we justify pumping into our atmosphere while we wait to get out of the “early days” of proof-of-work blockchains? How many people must be scammed for all they’re worth while technologists talk about just beginning to think about building safeguards into their platforms? How long must the laymen, who are so eagerly hustled into blockchain-based projects that promise to make them millionaires, be scolded as though it is their fault when they are scammed as if they should be capable of auditing smart contracts themselves?
The more you think about it, the more “it’s early days!” begins to sound like the desperate protestations of people with too much money sunk into a pyramid scheme, hoping they can bag a few more suckers and get out with their cash before the whole thing comes crashing down.