wg0

28 ms·

661.
by wg0 2y ago
On a tangent, this video[0] from Sabine Hossenfelder about academics in general is eye opening. In comments, veritasium[1] agrees: >After finishing my PhD I went to a university-led session on ‘What Comes Next.’ What I heard sounded a lo
662.
by wg0 2y ago
Training costing billions can't be. Biggest line item would be salaries. Probably. Also, revenue you can't take home. Doesn't matter how many billions. It is the profit that matters in the end.
663.
by wg0 2y ago
> OpenAI’s business model, as it exists today, doesn’t really inspire confidence, as it seems to exist on the promise of ‘jam tomorrow.’ Specifically, the firm has an income of approximately $3 billion per year, which is put in deep shad
664.
by wg0 2y ago
> The World Doesn’t Need More Features Whereas a PM's job is all about that. We have bloated products. Mostly. Feature packed with FOMO. Incoherent. Inconsistent. Incompatible. But features.
665.
by wg0 2y ago
For profit it will be when it will be profitable.
666.
by wg0 2y ago
Redis, Elasticsearch, Mongo and now WordPress - it seems that Open source is as good and only good when you and only you can sell it. The moment someone else starts to make money or more money then you could have off your effort, does thing
667.
by wg0 2y ago
Where are those Jobs that AI was supposed to take over? Especially programming?
668.
by wg0 2y ago
So product managers in these companies really think these products will make money and would become sustainable produc lone? Or it's just muscle flex and show off?
669.
by wg0 2y ago
Would be similar I guess? But those were contained islands not optimized for engagement.
670.
by wg0 2y ago
It seems Internet alone brought science and commerce but social media exposed and unleashed idiots.
671.
by wg0 2y ago
All the AAA games are just the same now. Essentially - almost same game mechanics with different 3d models and textures.
672.
Ask HN: Kodak, VW, Intel – Why big companies fail?
17 points by wg0 2y ago|38 comments
673.
by wg0 2y ago
colima + docker CLI goes a long way. $ colima start $ docker context use colima And that's it. And Kubernestes? No thank you, life is already hard as is.
674.
by wg0 2y ago
I believe that 99% of the time, a reuse of older technologies results in greater innovation that is more stable and a return is multiple of the underlying foundations. This applies here as well. Typescript could go a long way. But admittedl
675.
by wg0 2y ago
"Looks good on my CV" syndrome.
676.
by wg0 2y ago
One word comes to mind - "Juicero" [0]. https://en.wikipedia.org/wiki/Juicero
677.
by wg0 2y ago
Brutally minimized to first principles and to the very core. It's always about the journal and that's it. No fluff. Great project.
678.
by wg0 2y ago
Do we really know that these were explosives or lithium ion batteries on their own can be this dangerous?
679.
by wg0 2y ago
Now what is that? Didn't know about that.
680.
by wg0 2y ago
That reads like 99% cities are like San Francisco. The world unfortunately is lot more diverse than that.
681.
by wg0 2y ago
And those"Lean Startup - million copies sold" pivot gurus.
682.
by wg0 2y ago
>With sprints, there are no breaks, little autonomy, and insufficient time to prepare. That's the TLDR and on the dot.
683.
by wg0 2y ago
> Self driving works. No it doesn't. Only in San Francisco maybe.
684.
by wg0 2y ago
React?
685.
by wg0 2y ago
CoT again is result of computing probabilities on tokens which happen to be reasoning steps. So those are subject to the same limitations as LLMs themselves. And OpenAI knows this because exactly CoT output is the dataset that's needed
686.
by wg0 2y ago
True. It is not totally useless. There's a lot of utility.
687.
by wg0 2y ago
Because that's basic assumption. I'd want a discussion about academic papers from subject matter experts otherwise reading it aloud alone is not of any value let alone commentary on it with no understanding of it whatsoever.
688.
by wg0 2y ago
Is this written by ChatGPT? Feels like it. I think now even genuine content might feel the same.
689.
by wg0 2y ago
Not true. If the accuracy of human debate had this much room of error all the time when two subject matter experts are talking, we would not have the progress of civilisation that we have now. Room for error for sure is but at the very fron
690.
by wg0 2y ago
There's no guarantee that the discussion would be accurate. This stems from how the LLMs work.

More ›