Bryan Robinson

Bryan Robinson

@brob.dev

Developer, Educator, DevRel. I post about front end development, board games, content creation, and maybe the occasional philosophy.

823 followers439 following1.1K posts

I'm so tired of GenAI flyers. Not only are they all the same, they all suffer immensely from being way too busy, having really poor visual weighting of information, and an obvious inability to understand what's important information to get at a glance

Not to mention, watching the World Series on Fox and seeing the "Foxtoberfest" logo which 100% GenAI. Like 0 effort put in. You're a giant media conglomerate. Pay a freakin' designer.

So excited that @css-tricks.com is indie again. Already signed up for my yearly membership to get access to the Discord. I hope others will join the queue for that. Indie publications are the life blood of the type of internet that I want

Deep thoughts: Was the company Post Hog named after the logical fallacy Post Hoc Ergo Propter Hoc? It's a common fallacy when looking at data with poor attribution

Bryan Robinson reposted

This is exactly the principle that has fucked up everything from deli flyers to UX in billion-dollar products Once upon a time, even if you were doing something on your own, it took long enough to do that it made you think about what you were doing! You had time to think "hang on, this sucks."

I think the thing about AI and small business owners is that previously the design would have to be run past either a second person who know some amount of design, or they'd learn photoshop and some amount of design with it now they can ask the AI to do their bad design and the AI never says no

Anthropic released a blog post boasting their new version control system they created to handle 30k merges every day to their mono repo. With fewer than 3k devs on staff, I never want to hear about "The Human in the Middle" safeguard from them again. No chance code review functions properly there.

Bryan Robinson reposted

Some of you have been asking whether or not Vivaldi is affected by Chrome’s #AI API, which some websites are using for tracking purposes. The answer is straight-forward: no. ✌️ We disable all the Gemini-related code bundled in Chromium, effectively preventing websites from using this API in Vivaldi.

Anyone have recommendations for non-smart, wifi speakers? I've removed the last Echo Dot and miss conveniently streaming audio from my phone. No speech recognition, no AI, just relatively easy connection and mediocre sound quality (not an audiophile)

I'll start simply: There is no ethical use of LLMs. I mean that in both the same sentiment as "no ethical consumption under capitalism" but also deeper and more. The ills and harms are on the surface. The theft was blatant. The attempted wage theft was also blatant.

Andy Bell@bell.bz

There's a defeatist vibe rn and I think there's an inevitability feeling washing over us. There is no inevitability. If you're in a position to speak up against the direction the industry is taking, I'd say now is the time to do that. I certainly try because I cannot be fired and it does help!

Bryan Robinson reposted

Does my Speak-n-Spell feel pain when I type O-U-C-H? The jury is out. Philosophers and neuroscientists say that's fucking stupid, but I called a guy from the Center for Pretending that Household Electronics Are Sentient and he said there's no way to know for sure.

If you take the "AI" executives seriously, and you're a government official or world leader, your only recourse is to say "You have a month to cease operations. If you don't, you and your board and executive team will be charged with reckless endangerment"

Mind you, the way these execs talk IS ridiculous. The danger isn't AGI, the danger is the myriad ways the product is used, the environmental concerns, the psychological concerns, the extractive practices. But when they say these things to the UN, how do we not demand they shut down the company?

If you lost access to all LLM-based products today, you would be FINE. Perhaps that's the best reason to not use an LLM-based product today.

View full thread

If your team uses these products, one week a month, you have to do your job without them. We're always saying that a knowledgeable person needs to be controlling these tools. You need to know good work and review their output to ensure it's good. If you deskill, you'll be worse at that.

Therefor, one week a month, you need to be exercising your "non-LLM" skills. Make sure your knowledge doesn't atrophy. Make sure you can still be a good steward of your work. Don't build dependence.

If you lost access to all LLM-based products today, you would be FINE. Perhaps that's the best reason to not use an LLM-based product today.

Having regular discussions around the types of governance companies should have around LLM usage. I'm a "no use is ethical use" kinda guy, but if it's going to be used my favorite suggestion I've heard is that companies should enforce regular "detox" time periods.

I hadn't heard of @ellipsus.com before, but immediately signed up for it after watching this and exploring their stances

Adam Christopher@adamchristopher.me

Ellipsus is the best writing platform for a number of reasons, the most important of which is they are vehemently anti-AI, and their Emboss tool can prove you created what you did. And... THEY MADE A SHORT FILM ABOUT IT. Holy cats. www.youtube.com/watch?v=1iCI...

If you find the creation and continued development of a tool to be against your morals and ethics, how do you continue the usage of said tool? How do you "square that circle" as it were...?

Random realization: Losing the care for punctuation is the same as sending slop documents at work. You're moving the labor of understanding to the receiver. I have to read your sentence and still insert the pauses,etc. to make it make sense. You just passed the work on to me No AI More punctuation

How do people who advocate for expanding the use of LLMs then make jokes about how much LLMs hallucinate in certain tasks, and then continue to advocate for using them in those tasks with fewer and fewer guardrails?

Advocate: Let's use LLMs to analyze this data. * Advocate shares the results uncritically * Advocate: I used an LLM to read this other document and summarize. It hallucinated 5 times and made up dates. It's the worst. Ha ha ha. So silly. Advocate: Let's base decisions on what the LLM data analsys

One book a month for 5 years. Dolly's Imagination Library is one of the reason's my son is an avid reader. Every book, I'd start "<title> by <author>, sent to us by..." and I'd open the cover to where her picture was and he and I would say together "DOLLY!" Thanks for that and much more Dolly.

Bryan Robinson reposted

CAPITAL: We have invented a machine that can cross picket lines PROGRAMMERS: I bet this will give us four day weekends. The left need to get behind this. What else can it do CAPITAL: Psychosis! Deepfakes! Propaganda! Plagerism! Yaml! Todo Apps! PROGRAMMERS: Run those last two past me again

Bryan Robinson reposted

if open source has proven anything at all it's that you can have a project with hundreds of brilliant technical contributors and not one of them will be even remotely capable of designing a decent user interface

winter@madoka.systems

at least on this project, you're telling me out of the 50+ contributors, not one of them wants to design the site or write a fucking readme? shame on you

Why did social media decide that if I turn on the audio for one video post, all the video posts for that session should default to audio on? I don't scroll them often, but both Facebook and LinkedIn do that. It's an odd choice

If no one holds slop accountable, how do we stop people from creating slop?

Why is it that the platitude "Humans will own the output" in company AI marketing releases never actually describe what that means? If you make something that does something awful (customer data breach) that's easy to punish... but what if you just make your coworkers lives all slightly worse?

Why is it that the platitude "Humans will own the output" in company AI marketing releases never actually describe what that means? If you make something that does something awful (customer data breach) that's easy to punish... but what if you just make your coworkers lives all slightly worse?

"The poet's job is to notice" quotes the footer of the site in GIANT letters. The best way to train yourself to notice is to offload your noticing and creativity, it would appear

Write your own poetry. Even if it's crap. It'll be better than whatever this is

Bryan Robinson reposted

📣 I’m excited to announce!! 📣 I’m partnering with Unstable Games to start a sister studio to them named… ✨ Ordinary Magic!! ✨ I’m so excited to bring 3 BRAND NEW GAMES under this new studio to you this Fall!!! 🍁

gamefound.comOrdinary Magic by Unstable Games - Gamefound3 New Games, 1 Extraordinary World. Explore this visionary studio partnership between award-winning game designer Marceline Leiman and Unstable Games.

Totally forgot about the time that I went on local news in 2020 to talk about the ridiculousness of speed ticket cameras and I drop language that's ALMOST "human in the loop" before I even thought I'd need to know that phrase. www.actionnews5.com/2020/02/27/i...

actionnews5.comThe Investigators uncover flaw with school zone camerasA school zone camera erroneously issued speeding tickets to at least two drivers, and three people who reviewed those tickets didn’t catch the mistakes.

So, let me get this straight. Instead of curtailing expansion of a ridiculously wasteful set of technologies, instead, let's give them carte blanche to be as wasteful as possible. Excellent. No notes

In this era, the tech industry is just disappointment after disappointment. I don't think there's anything left that's truly worthwhile

Let's say you have a personal bot that saves you 5 minutes a day... Let's say it costs you an average of $3/day. That's $1095/year. That's mind-bogglingly bad economics, if you ask me. But you didn't ask me, so I'll just be quiet over here

Thinking about the next phase of my play library... As a proof of concept, I'm thinking of opening up free checkout of my library to my immediate neighborhood to see how things are before charging money for membership. Would you check out games from your neighbor?

I keep hearing different use cases that "we could use AI" for. So far, EVERY TIME, I think "We have been able to do that use case for a decade without AI with lower incidence of incorrect data, lower cost, and higher reliability." EVERY. TIME.

Bryan Robinson reposted

Today the good people of the world commemorate the passing of Google Reader thirteen short years ago on July 1, 2013. To honor their memory (and in lieu of flowers) please add an RSS/Atom/JSON feed to your web site and keep building a web that connects people in meaningful and positive ways.

Ownership of media is so important. We get tricked by convenience into giving away our ability to actually own any sort of physical media... Sony announcing the death of game discs is just another step on a long street to remove agency and control over what we "own"