likium

7 ms·

31.
by likium 1y ago
Unfortunately, most people just want eggs, not the burden of actually owning the goose.
32.
by likium 1y ago
Citing is not referencing links, it's finding primary sources and fact-checking. Kurzgesagt[1], an informational YouTube channel, has had issues with LLMs citing LLM generated content. [1]: https://youtube.com/watch?v=_
33.
by likium 1y ago
If customer data is considered edge, then what’s core?
34.
by likium 1y ago
What's the benefit of using K3 on a single node?
35.
by likium 1y ago
They're called search-based targeted ads. That's where Google's AdWords started from, and you can still do that if you wish. But personalized ads work a magnitude better.
36.
by likium 1y ago
The incentive to show post-purchase ads is: 1. people didn't like nth purchase so they're looking for a substitute. 2. reinforce purchase behaviors / reduce buyer's remorse. Ads work at human psychology; they're not
37.
by likium 1y ago
We need lower barriers so users aren’t beholden to a handful of walled gardens. When big tech can’t collude, salaries might in fact go up.
38.
by likium 1y ago
Agreed but hooks did solve composability. It’s easier to implement very complex handling, like drag and drop interactions.
39.
by likium 1y ago
My point is humans make mistakes too, and we trust them, not because we inspect everything they say or do, but from how society is set up. I'm not sure how up to date you are but most AIs with tool calling can do math. Image generation
40.
by likium 1y ago
I do believe that's where we're heading, people holding jobs to hold accountability for AI.
41.
by likium 1y ago
We place plenty of trust with strangers to do their jobs to keep society going. What’s their error rate? It all ends up with the track record, perception and experience of the LLMs. Kinda like self-driving cars.
42.
by likium 1y ago
A better analogy would be musicians. Recorded music is around but some musicians still make a living, mostly off live concerts and merch. But it might also go the way of pottery, glass-making and weaving. They’re still around but extremely
43.
by likium 1y ago
Agreed. In that world, fame and power becomes more important since wealth no longer matters.
44.
by likium 1y ago
That's what Claude Pro will give you... now. As with Cursor there's no guarantee that it'll last. Next year, next month, next week, they may change their pricing when the competition runs dry.
45.
by likium 1y ago
What tool do you use?
46.
by likium 1y ago
We only have access to local models because they're subsidized too. There's nothing to prevent companies or state actors from paying/funding for increased probability and inherent bias. Also local models are close in capabili
47.
by likium 1y ago
Notably, this is written by David Singleton, who at the time was the CTO of Stripe.
48.
by likium 1y ago
If your company has a good component library, you can really stay in that circle for 95% of the time. If you need that 5%, usually there's a library unless you're doing something really bespoke like canvas/webgl. And AI reall
49.
by likium 1y ago
There is a viable market, it exists as a word + excel combination like Notion, Coda and Monday. Notion just recently had an email and calendar client. And I don't think non-US or non EU-wages, or being open source would help. Microsoft
50.
by likium 1y ago
I’m making a presentation ai assistant. Existing ones don’t cut it… they’re really nice to demo but not good enough for professional use. Full MVP by next week. https://unblank.ai
51.
by likium 2y ago
Building a multi-modal llm app across multiple platforms and syncs. Doing this is both harder and easier than I expected. Can’t wait to ship it out into the real world soon.
52.
by likium 2y ago
I’m concerned. Not just privacy but AI request forgery will also be a thing. Running local models don’t protect you from prompt injection attacks or hallucinations. There are some startups building capabilities apis to limit that but most
53.
by likium 2y ago
The places where automation is most needed are for non-technical folks. To “write a script” is a huge hurdle.
54.
by likium 2y ago
> You earn the right to take on bigger challenges and no one in this space has earned the right yet including us. Not entirely. We're in interesting times where products with better models can suddenly leapfrog and displace even cur
55.
by likium 2y ago
Nice playground! I built an interactive guide a few years ago with a similar playground at the end: https://lik.ai/guides/an-interactive-guide-to-flexbox/ It was useful to refresh my memory here and there with it.
56.
by likium 2y ago
Just curious if the processes got tuned/adjusted as a result? And what were the negative side effects?
57.
by likium 2y ago
Much could be said about open source libraries that companies release for free to use (kubernetes, react, firecracker, etc). It might be strategically make sense for them so in the meantime we’ll just reap the benefits.
58.
by likium 2y ago
Starlark is widely used by Bazel and Buck. That would fulfill similar properties and build system folks have one fewer thing to learn.
59.
by likium 2y ago
Won't the right abstractions like NumPy make it possible for researchers to obtain generally performant code, and lower the bar to write specialized, optimized code without having to drop all the way down to CUDA?
60.
by likium 2y ago
It’s hard but Stripe’s design system does font metric calculation and adjustments. Haven’t seen a misaligned icon or button in a while.

More ›