I was told that calling some people at YouTube "incompetent" in my article was "disrespectful", so I want to clarify something because it's important.
This industry already has way too many YouTube yes men who will defend almost anything the platform does.
I'm not interested in being one of them.
I also know that some very important people in this industry read what I write (including people at YouTube). We talk behind the scenes, we debate, disagree on a lot of things so I'm fully aware that my words have an impact.
I see this as a responsibility. I have technical knowledge in areas most creators don't, so I can see problems in these systems that the average creator isn't equipped to spot.
If I think something is good I'll say it. If I think something is wrong I'll say it too, but I'll explain why. Constructive criticism is how things improve.
But when I've been criticizing some of the same problems for years and nothing changes, creators also deserve to hear a different voice than the usual "YouTube knows what they're doing".
Especially when these decisions have real consequences.
I've been a content creator for a good portion of my life and have spent enough time around creators to know how deeply these numbers affect them.
I knew a fellow creator who took his own life.
I'm obviously not saying YouTube Analytics caused that (that would be absurd and unfair). His personal life was a mess, but his channel was also his full time job and its poor performance at the time was one of the many factors that were weighing on him.
So when YouTube designs analytics that are read by millions of creators every day and explicitly teaches them wrong things about how those numbers should be interpreted, I think those decisions deserve very serious public criticism.
And if after looking at the reasoning behind one of those decisions I conclude that it is incompetent, then I'm going to call it incompetent because that's what it is.
The standard should be extremely high and the criticism should match the scale of that influence.
That's what I'm doing.
@CyrilleDaclinat J'ai l'impression de savoir ce que tu vas dire dans la vidéo, pas de gap qui me pousse a cliquer. Je me dis ok donc il va me dire un truc du genre "La structure = Hook + xyz+ climax"
Je trouve la compo de la thumbnail super bien executée, mais elle en dit trop.
Introducing Click IQ
A 6th sense some of you have but don't know yet:
- the ability to feel what millions will click before they do
Find out if you have it, it's free.
Alors, bulle IA ou pas bulle ?
A mon sens :
1/ si l’information est aussi accessible aux particuliers, dans les médias mainstream, c’est que ça n’est pas une information financière rare.
Donc que c’est probablement une info gratte-buzz, un leurre, ou juste une erreur.
2/ on ne regarde pas au bon endroit économiquement parlant.
2023 et le raz de marée ChatGPT ont probablement sauvé une finance mondiale qui cherchait sa martingale pour maintenir une croissance continue qu’il ne fallait surtout pas stopper.
J’ai eu l’opportunité de développer ce sujet avec @RebaudThomas récemment justement !
Parce qu’il y a quelque mois, @momaurane m’a contacté pour me proposer de devenir celui qui embêterai Thomas à chaque épisode de leur nouveau podcast sur la Tech.
J’ai accepté le job avec grand plaisir 🫡🤓🤭
Je vous mets le lien de la chaîne YouTube ici : https://t.co/kVqwDO1BGr
Bon visionnage !
Hey ! I’m building AI agents that run top-tier YouTubers' operations, starting with the 3 most important things: the idea, the title, the thumbnail.
We built a proprietary memory layer with semantic embeddings that compounds with every video the creator publishes, every comment from their audience. All the agents read from it and write back to it to get smarter and sharper.
Each agent also comes with its own native UX, optimised to capture the creator's taste as structured signal, instead of relying on low-bandwidth chat-bot-style text-based inputs.
I’m building AI agents that run top-tier YouTubers' operations, starting with the 3 most important things: the idea, the title, the thumbnail.
We built a proprietary memory layer with semantic embeddings that compounds with every video the creator publishes, every comment from their audience. All the agents read from it and write back to it to get smarter and sharper.
Each agent also comes with its own native UX, optimised to capture the creator's taste as structured signal, instead of relying on low-bandwidth chat-bot-style text-based inputs.
I get it. One can easily argue it’s better to do nothing than an inauthentic/weak goal.
But i really think its wrong way to look at it
same goal can feel inauthentic in one creator’s video but perfectly understandable in the other
it’s never the goal in itself
The era of inauthentic goal-setting in entertainment based YouTube creators is fascinating to see.
E.g: Some type of crossing America video:
“And I make this video to bring a cake to my friend from high school I haven’t seen in years!”
Not because it’s the actual reason, but because it’s a very thin, shaky layer of an excuse to try to make the viewer care about the goal.
How do you battle this as a creator?
Writing a newsletter on this tomorrow.
I think I just finished vibe-coding the best AI thumbnail tool on the market. Give me any thumbnail to create or edit (with the tiniest details) and I’ll do it for you to demo.
@CyrilleDaclinat Idée comme ça pour la minia jouer plutôt sur les animaux que sur Darwin pour illustrer la loi du plus fort. À la manière de la trend « why it sucks to be born as » par exemple jouer sur un grand prédateur en mode « looser »
@CyrilleDaclinat Je dirais :
- on voit pas assez le côté humour dans le packaging, ça peut rendre la vidéo unique, là c vu et revu
- c’est pas assez “noir ou blanc” dans le choix des mots, pas assez extreme comme pov. Debunk c’est gentil, le “gros mensonge” le “scam” “l’escroc” c’est incisif