There’s a lot of irony to spammer ads that (mis)use Elon’s picture getting through the programmatic approval processes.
220
Vox just released a cool primer on cement and carbon impact.
youtu.be/asLWBGtAhZk
🔮 #coolcompany
Cement is one of the most abundant man-made materials. Unfortunately, the production of cement causes 8% of the global emissions.
@SublimeSystems is electrifying the production process, significantly reducing the missions -- with a goal to going to zero.
2
246
But have you tried an original Frappuccino while in Eataly?
You’re unable to view this Post because this account owner limits who can view their Posts. Learn more
1
170
Yeah sure, generative code is great, but have you tried running PROD on it?
It is absolutely wild watching people who are super senior and respected in our industry capitulating to the AI hype. I get it- it’s super super exciting. But even if AI ends up changing a lot, you can’t just throw away decades of programming fundamentals.
1
141
😟 nos.nl/l/2468298
Kom er maar in @FilemonW ?
59
The thing I least get about the imperial measurement system, being based on very concrete units of far gone centuries (“stones”) is why they didn’t call it Fahrenheat.
111
Reminder voor alle OV vakbonden dat reizigers geen partij zijn. Als je zo kort vooraf kunt besluiten te staken, kun je dat ook afblazen als dat onverantwoord blijkt met nieuwe info. ⛄️ Ik heb er meerdere zien gaan 🚲 jullie cadeau aan hen is weken gips. @FNV
109
Unfinished thought. Super interesting ChatGPT is still solving IR at its core.
It won’t kill Google: needs efficient indexing to stay UTD. So even if its better at locating knowledge and understanding intent than (the many pieces that are) Google search…
1
1
2/2 … if we assume Google has a nearly equivalent LLM… (which is trained on that index!), it has 2/2 of the core components necessary to replace itself vs only 1/2 for OpenAI (with a low likelihood OpenAI has built an equivalent indexing engine because of the network).
1
1
Kwam gisteren half wegdommelend langs @Jinek_RTL, waar ik ChatGPT code zag “schrijven” op een prompt van @AlexanderNL.
Het lijkt heel wat, maar het is code *nadoen*, niet schrijven.
Alas, het staat letterlijk in de examples
github.com/Uberi/speech_reco…
197
ChatGP will bring us into the golden age of actual, human fact checking.
Extremely ironic the surge in information it is trained on and all the issues with truthfulness become transparently visible once you remove the authority derived from humans that generated the knowledge.
1
2










