If your primary source for announcing upcoming features and behind-the-scenes content is an X-Twitter account, and not an RSS feed you own, you're doing it wrong.
Note that this does not mean you can not also use your account(s) of choice as secondary channels for communication efforts.
If you are a browser company claiming to inspirit what the internet can be, I'll be expecting more.
Just my thought of the day.
Per Axbom
@mjj It's an interesting point. The way it works Signal doesn't store any information about its users nor any chats... they're all on the user's device. As long as one trusts this, it doesn't matter if a government can request data, Signal doesn't have it.
I'm not as sympathetic to teachers worrying about AI-written assignments as I am to students worrying that their teachers won't be able to help them develop a capacity for AI appraisal.
"Oh yeah, I really love this website!"
Humans are fascinating when it comes to language. For example, the above statement can be expressed, using voice (and to an extent also body language and context) so that the person hearing the statement clearly understands that the person behind the statement actually means the polar opposite.
Now imagine using a language model to transcribe and provide a summary of an interview. Imagine how many human nuances will get lost. And how many faulty ideas may be invented and adopted because people believe a language model is "smart", even when its output is untruthful or random.
No, I would never use a language model in UX work. I could never assume and feel certain that I am able to find and correct all the mistakes made by a machine.
#languagemodel #AIHype
1999 skrev vännerna Levine, Locke, Searls och Weinberger ett manifest: Cluetrain Manifesto. Det var ett försök att skapa förståelse för kraften i nätverkskoppplade marknader och hur företag måste förhålla sig till att människors sätt att föra samtal om dem.
Vilken oerhörd makt det kan ge när människor samtalar med varandra OM företag och organisationer, varnar och tipsar, och även pratar direkt med anställda på INSIDAN. Via internet förstås.
Jag tror jag känner många som kan minnas detta manifest, som bestod av 95 teser (eller "ledtrådar"). Om än lite vagt. Det här var en tid när ett betydande antal av oss hade en naiv tro på internet som en ostoppbar demokratiserande kraft.
Cluetrain Manifesto var också en bok, men själva manifestet publicerades öppet och publikt, och översattes till många språk. Bland annat översattes det till svenska av mig.
När två av originalförfattarna valde att 2015 följa upp med "nya ledtrådar" fick jag äran att översätta även dessa till svenska.
I de nya teserna pekar Doc Searls och David Weinberger på hur internet är hotat, och att vi måste agera. Det är hotat av de stora plattformarna och deras makt att styra vad vi ser och inte ser, vem som har tillgång och inte. Även om det ursprungliga manifestets teser fortfarande är relevanta så blir 2015 års teser ett erkännande av hur man missade att förutse de stora drakarnas övertagande.
Men det är också en kärleksförklaring till webben. Och en påminnelse om att vi inte är klara. Aldrig blir det. Att vi själva kan välja hur vi agerar och vem vi stöttar. Att vi fortfarande har den där makten som uttrycktes i manifestet från 1999.
Jag läser om alla dessa teser ungefär vartannat år. Det är oerhört nördigt, men också en fin återblick och påminnelse om förväntningarna som fanns långt innan de fåtal plattformar vi agerar på idag etablerade sig.
Framför allt är det så många tankar och spår att de garanterat ger insikter och inspiration till nya tankar. Det är nog det jag gillar mest. Att de uppmanar mig att tänka annorlunda, när jag riskerar att acceptera för mycket av nuläget.
För mig känns det som att 2023 står för något nytt. Ja, AI har sin hajp, men det är också väldigt stökigt på de där plattformarna. Fler makthavare i politiken har en större förståelse för vad internet är. Det finns en öppning att återta lite av den där makten – att fler ska få äga sitt innehåll. Att fler ska återta förmågan att kommunicera och konsumera innehåll på sina egna villkor. Att fler ska få äga det de köper (i stället för att digitala filmer och böcker vi "köper" i praktiken bara hyrs).
Skulle vara oerhört roligt att höra vad du tycker, om du läst dem. Och om du alls kommer ihåg dem.
1) Cluetrain Manifesto i svensk översättning:
https://axbom.se/cluetrainmanifesto/
2) Cluetrain Manifesto i original (1999):
3) New Clues (nya ledtrådar) i svensk översättning:
https://axbom.se/cluetrain-nya/
4) New Clues i original (2015):
@Loukas
Friend of a friend of the publisher? 😉
Do you find it surprising that this is published in Time?
Judy Estrin is on fire in her op-ed for Time Magazine.
“What is happening is not just a battle for market control. A small number of tech titans are busy designing our collective future, presenting their societal vision, and specific beliefs about our humanity, as theonly possible path. Hiding behind an illusion of natural market forces, they are harnessing their wealth and influence to shape not just productization and implementation of AI technology, but also the research.“
She goes on to talk about the politics of inevitability and how we are tricked into thinking the future dictated by big tech is unavoidable. (Pro tip: it isn’t)
But this sentence actually caught me off-guard: “On the current trajectory we may not even have the option to weigh in on who gets to decide what is in our best interest.“
Do read it.
https://time.com/6302761/ai-risks-autonomy/
#DigitalEthics
From my talk on AI and Human Rights in January of 2022…
A tiny, tiny proportion of humanity (3 million AI engineers is 0.0004% of the world's population) has the knowledge and power to build machines that are "intelligent" enough to potentially decide who gets a job, who can obtain financial aid, who the courts trust, or whose DNA and biometric data will be harvested by marketers.
We have a technology in the hands of a few, born in societies of abundance, and powerful enough to shape the outcome of lives across the planet.
This asymmetry of knowledge and power is a core challenge for human rights. An algorithm that is oppressive, racist, bigoted, misogynous and unfair can be all those things more efficiently than any human.
All while maintaining an aura of "neutrality" and superior smartness.
We are creating a more efficient version of the past instead of reimagining the future we want.
There is something to be said for human inefficiency, the ability to pause and reflect, taking time to question and taking time to listen. Do we want to allow AI to erase the language, alphabets, culture and stories of poorer countries?
We need to figure out together what a more equal and peaceful world means. Only after that we discuss how AI can help us achieve the positive technical and ethical outcomes that will lead to that future.
Instead of shipping AI across the world we could be engaging in inclusive, pluriversal conversation about what that world should look like. What we keep and what we remove. Whose voices matter.
I believe in AI. I believe in its enormous power to assist understanding. I am not interested in stopping progress. I am urging all of us to to reconsider how progress is defined, and how it is distributed.
Video, slidedeck and transcript:
https://axbom.com/aihr/
@AnnemarieBridy
Absolutely, and it’s positive that they strive for more.
For me the dilemma starts before the LLMs, as the amount of content available online for training large language models in different languages is quite unbalanced, as the article suggests. Hence their performance will differ significantly depending on language. It’s a bias built into the fabric of the internet (and hence the training data) rather than an inherent problem of language models.
We often forget that there are more than 6,000 languages in the world. I take every opportunity to remind myself of this. 😊
Innovation ≠ Progress
Some innovation will improve lives.
Some innovation will worsen lives.
Some innovation will help one group of people while creating struggles for other groups.
Striving to increase innovation does not guarantee a net increase in wellbeing.
Progress requires a moral stance that continuously considers and is explicit about how the impact of an action helps someone without harming someone else.
And here is the twist:
Progress does not even need innovation.
The primary ingredient of progress is something we all already carry with us.
It's kindness.
And if you find yourself on a path where kindness is missing, the path doesn't define you. You can forgive yourself and choose another path.
This is saying: You must first be kind to yourself.
Helping others forgive themselves.
Helping others towards paths where they are equipped to practice kindness.
Now that's innovation.
None of this is easy. None of this is a straight line. But sometimes just a simple reminder.
A reminder in the right moment.
Can trigger progress.
"Our data shows that a little more than half the sites on the web use English as their primary language. That’s a lot more than one might expect, given that native English speakers only make up just under 5% of the global population."
"Millions of non-native English speakers and non-English speakers are stuck using the web in a language other than the one they were born into. And since publicly available text on the internet is now often being used to train large language models like Bard and GPT-4, it suggests we’re already building the same imbalance into technology’s next frontier: artificial intelligence."
https://restofworld.org/2023/internet-most-used-languages/
#DigitalEthics
At a café working on my Macbook and have my iPad set up to the right of my computer. I knew I could use it as an extension of my desktop, but I actually had no idea I can just pull the cursor over and control everything on my iPad from my laptop. Wild.
And did the computer somehow figure out that my iPad is on the right by itself? I don't remember ever telling it...
First day back from vacation and I'm feeling like a complete novice. 😅
These numbers on how we best learn keep being quoted but there is unfortunately very little substance to them.
Read this for a reasoned perspective on the numbers (originally published in 2002):
@eternaltyro That was a really good read! Thank you 🙏
Appreciated the nuclear vs solar narrative, which brought a lot of clarity.
I haven’t seen Oppenheimer yet but it feels a good opportunity today to draw these parallells as this is currently top-of-mind in more societal contexts. This has given me lots of ideas for new content.
@chris Gah. I actually did not know that.
@potato_lisper I agree, this is a disadvantage. But less of a dilemma in situations where everyone’s phone number should already be known by all participants.
If you are seriously considering new software for video meetings, don’t forget about Signal.
Signal currently allows up to 40 people in the same meeting. The part that perhaps isn’t intuitive is that people have first be added to the same group chat, and then the video meeting can start from there.
Using Signal Desktop, you can share a screen or separate window while on a call. You can share your screen if no one else is presenting. Your own video will not be shown while you are presenting.
To start or stop sharing while on a call, select the presenting icon with the box with an arrow pointing up.
Why Signal? In their own words:
”State-of-the-art end-to-end encryption (powered by the open source Signal Protocol) keeps your conversations secure. We can't read your messages or listen to your calls, and no one else can either. Privacy isn’t an optional mode — it’s just the way that Signal works. Every message, every call, every time.”
My advice? Always use Signal [1] (or equally secure software) for internal meetings with up to 40 people. And also suggest it, when it makes sense, to external suppliers and clients. And for education/webinar/all-hands contexts, consider setting up your own server with Big Blue Button [2].
#VideoMeetings #VideoConferencing