Statement on Generative AI | Ben Myers
I endorse this statement.
I endorse this statement.
I want to live in a future where Artificial Intelligences can relieve humans of the drudgery of labour. But I don’t want to live in a future which is built by ripping-off people against their will.
- Be skeptical of PR hype
- Question the training data
- Evaluate the model
- Consider downstream harms
There’s a general consensus that large language models are going to get better and better. But what if this as good as it gets …before the snake eats its own tail?
The tails of the original content distribution disappear. Within a few generations, text becomes garbage, as Gaussian distributions converge and may even become delta functions. We call this effect model collapse.
Just as we’ve strewn the oceans with plastic trash and filled the atmosphere with carbon dioxide, so we’re about to fill the Internet with blah. This will make it harder to train newer models by scraping the web, giving an advantage to firms which already did that, or which control access to human interfaces at scale.
Design systems as codified constraints.
Everything old is new again:
In our current “information age,” or so the story goes, we suffer in new and unique ways.
But the idea that modern life, and particularly modern technology, harms as well as helps, is deeply embedded in Western culture: In fact, the Victorians diagnosed very similar problems in their own society.
Like Bastian, I’m making a concerted effort now to fly less—offsetting the flights I do take—and to take the train instead. Here’s a description of a train journey to Nottingham for New Adventures, all the way from Germany.
If you treat data as a constraint in your design and development process, you’ll likely be able to brainstorm a large number of different ways to keep data usage to a minimum while still providing an excellent experience. Doing less doesn’t mean it has to feel broken.
The cosmonaut counterparts of the Mercury women astronauts: Zhanna Yorkina, Irina Solovyova, Tatyana Kuznetsova, Valentina Ponomareva, and Valentina Tereshkova.
Ponomareva recalled there being no envy between the women in the squad. According to her, it was a healthy spirit of competition. Everyone did their best to be number one, but also supported each other’s efforts.
One of those cosmonauts went to space: none of the women training for the Mercury missions did. There would be a shockingly gap of twenty years between the launch of Valentina Tereshkova and the launch of Sally Ride.
Designing your design process:
- Know your strengths and focus resources on your weaknesses.
- Learn to identify the immovable objects.
- What has to be perfect now and what can be fixed later?
The slides from a presentation by Drew on all the functionality that browsers give us for free when it comes to validating form inputs.
Half the battle of the web platform is knowing what technology is out there, ready to use. We’re all familiar with the ability to declare validation constraints in our HTML5 forms, but were you aware there’s a JavaScript API that goes along with it?
An astoundingly great piece of writing from Paul Ford, comparing the dot-com bubble and the current blockchain bubble. This resonates so hard:
I knew I was supposed to have an opinion on how the web and the capital markets interacted, but I just wanted to write stuff and put it online. Or to talk about web standards—those documents, crafted by committees at the World Wide Web consortium, that defined the contract between a web browser and a web server, outlining how HTML would work. These standards didn’t define just software, but also culture; this was the raw material of human interaction.
And, damn, if this isn’t the best description the post-bubble web:
Heat and light returned. And bit by bit, the software industry insinuated itself into every aspect of global enterprise. Mobile happened, social networks exploded, jobs returned, and coding schools popped up to convert humans into programmers and feed them to the champing maw of commerce. The abstractions I loved became industries.
Oof! That isn’t even the final gut punch. This is:
Here’s what I finally figured out, 25 years in: What Silicon Valley loves most isn’t the products, or the platforms underneath them, but markets.
This JavaScript training course looks like it’s really well planned to take you from zero to hero—there’s a whole module on progressive enhancement.
A social network for snippets of JavaScript effects in canvas, written in 140 characters or fewer. Impressive!
Hit this URL to give yourself a design constraint (or obstruction). Kind of like Brian Eno’s oblique strategies but with different categories of constraints: formal, methodological, and conceptual.
JP Rangaswami also examines the rise of the platforms but he’s got some ideas for a more sustainable future:
A part of me wants to evoke Jane Jacobs and Christopher Alexander when it comes to building sustainable platforms. The platform “community” needs to be cared for and looked after, the living spaces they inhabit need to be designed to last. Multipurpose rather than monoculture, diverse rather than homogeneous . Prior industrial models where entire communities would rely on a single industry need to be learnt from and avoided. We shouldn’t be building the rust belts of the future. We should be looking for the death and life of great platforms, for a pattern language for sustainable platforms.
I frequently see web developers struggling to become better, but without a path or any indication of clear direction. This repository is an attempt to sharing my experiences, and any contributions, that can help provide such a direction.
It’s broken down into four parts:
I don’t necessarily agree with everything here (and I really don’t like the “rockstar” labelling), but that’s okay:
Anything written here is open to debate and challenges are encouraged.
I love, love, *love, traintimes.org.uk—partly because it’s so useful, but also because it’s so fast. I know public transport is the clichéd use-case when it comes to talking about web performance, but in this case it’s genuine: I use the site on trains and in airports.
Matthew gives a blow-by-blow account of the performance optimisations he’s made for the site, including a service worker. The whole thing is a masterclass in performance and progressive enhancement. I’m so glad he took the time to share this!
A write-up of the BrightSparks programme that Clearleft is taking part in.
Each company agreed to help support one local child from a low-income family, on free school meals or with a yearly household income of under £25k.
An online training course that will banish all fear of the command line, expertly delivered by the one and only Remy Sharp.
For designers, new developers, UX, UI, product owners and anyone who’s been asked to “just open the terminal”.
Take advantage of the special launch price—there are some serious price reductions there.