> That's full multi-modal training with embodied agents (aka robots). 1x, Figure, Physical Intelligence, Tesla are all making rapid progress on functionality which is definitely beyond frontier LLMs because it is distinctly different.
Cool, but we already have robots doing this in 2d space (aka self driving cars) that struggle not to kill people. How is adding a third dimension going to help? People are just refusing to accept the fact that machine learning is not intelligence.
FYI, Python is named after Monty Python , who got into all kinds of trouble making fun of and saying things that upset basically everyone (religion to LGBT), which is ironic considering the circumstances.
I seem to recall reading articles like this in 1997 about the internet.
“What’s there to do really?” they asked. Whatever annoyance somebody had with the state of the early web was magnified into a portrait of decline.
It was only for nerds. Or maybe too shallow with ugly amateur content. Or too commercialized already. Or maybe it was never going to be a successful platform for business because nobody is crazy enough to put their credit card number in an online form. Etc.
These contradictory complaints were all present back in the day, but in retrospect there was so much room for both technical and social growth that they just seem quaint now.
Unless you're working on undergraduate integrals then it's pretty useless for advanced stuff without writing your own library. By that point you may as well write it in a language that's more performant and cheaper.
At first, I thought you built a justwatch.com clone. But, after a quick inspection I realized you solve a related but different use-case where the user already paid for the streaming service holding the rights to the media, but not necessarily match the user’s country location.
By enabling the user to check in what countries the media is available you also unlock a new business model based on VPN affiliation programs (brilliant!) instead of streaming service affiliation programs like justwatch does. I wonder if both revenues models are complementary in the long run.
It’s nuts that watching media from another country for a streaming service I pay for is as easy as changing my IP to the desired country. Does it really work in this way?
Also, I hope you can provide some light on whether you are gaining any traction and unlocking vpn subcriptions revenue.
Suggestion: let the user select from what country it connects And/or automatically detect it. I’m in Europe and it’s confusing it offers me to stream from my own country using a vpn.
One of the big challenges is being able to frame real world concrete problems in terms that are solvable by quantum algorithms. There are a shocking lack of examples out there.
Even when the hardware is there, it isn’t clear how quantum computing can do useful things. But I suspect this is a solvable problem of good communications. But right now, experts in quantum don’t seem to be able to provide these examples. Or they don’t exist.
What is the average ROI on advertising by using google AdSense et. al?
Advertising is an assumed cost of business for any medium to large company these days, but I feel there's a growing silent minority who slowly believe, slowly know, that advertising in the online space is unfeasible for a majority of products. Advertisers like google make so much money because online advertising is one of those things that is super easy to keep throwing more money at. I wonder how many companies would benefit from entirely ditching online advertising and pursue alternate methods like direct outreach or sponsored content.
We might see 4,000-10,000 advertisements a day[1]. Do you remember two advertisements you saw today? Mere-exposure effect is the only thing going for it. Which may be fine if you're Coca-Cola, not so much if you run an online clothing store.
The answer was there all along since, at least Maven (but probably earlier). When specifying dependencies in Maven, you need to also provide checksums. Really hard to screw that accidentally.
But, the real problem is that any such system needs moderators. Verifying submissions is a tedious and difficult task with a lot of responsibility, if we are talking about something the size of PyPI. If they want to properly process the volume of submissions they have today, they need an army...
On the other hand, 99% of all stuff stored on PyPI is absolute junk. Yet on the other hand, people are used to there being dozens of versions of every package and packages declare dependencies very liberally.
All leads to the situation where there's a "frontier" of packages that can be installed together, but anything that lags 3-4+ versions behind the frontier is just taking space. Even if it's not malicious, it's not installable anymore because there's no more support for the platform it was written for.
But this is not the end of it. There are tons of simply broken packages, where some part of the archive is missing, or is malformed etc.
Proper moderation would have you submit your code, then review it, then publish it. Possibly, rejecting it in between if the code doesn't match requirements. Most people who publish their stuff to PyPI don't understand what needs to be in the package. Even packages like NumPy are full of useless junk.
But this will never happen, because the "community" grew used to the circus PyPI is. Python community doesn't care about the quality of the code, safety or anything other than a very shortsighted "make it work now" kind of goal.
There won't be a revision that purges junk or makes adding more junk harder. There will be another "quick fix", that will be obviously bad, but will kick the can down the road for a while longer.
It's a broken argument with some truth too it. IE you can run a SQL query put the result in a dataframe to dump it to an interchange format. But in the same breathe... If you learn to use SQL a great deal of workloads often used via dataframes APIs kind of disappear, and in doing so learning a new language to use a new dataframes API isn't really worth it.
It’s the default substitution done on iOS devices - the OS used on the most widely used smartphone in the US - this is a lot more common among HN users than copying and pasting out of MS Word??
> (disclaimer: lead brand manager, been at the table of a lot of the meetings where these logo changes happen.)
This is not a disclaimer. This is, in fact, the opposite of a disclaimer. It is a "claimer", if you will, claiming experience and authority on the topic.
A disclaimer is when you state a reason why people might want to discount your opinion, usually due to a conflict of interest. E.g., "I work at Revolut."
"We have also benefited from the various open-source packages on which this
textbook depends (see appendix G). The typesetting of the code was done with
the help of pythontex, which is maintained by Geoffrey Poore. The typeface used
for the algorithms is JuliaMono (github.com/cormullion/juliamono). The plotting
was handled by pgfplots, which is maintained by Christian Feuersänger"
It’s true in the US too, depends on the risk of terminated employees causing intentional harm. Most of the time layoffs come with severance making it similar to your “garden leave” situations.
People act like this is some spiteful thing he's doing in order to just post edgy memes or have a 'private' social media for himself.
Twitter, despite being a toxic place the majority of people avoid, brought in over 5 billion dollars last year. If elon removes bots, welcomes non-extremists back on, gets comedians and entertaining accounts back on board and lets people say what they want instead a bot army of shills repeating verbatim over and over and over... he could see that revenue rise quite a bit through people actually seeing value in advertising on twitter again. If he can keep operating costs down and get the ad revenue up further, he'll be repaying his initial investment within a few years, and if he takes this private he can IPO it again or sell it to someone else privately.
I wish him the best and hope he truly makes twitter somewhere that you visit that isn't just rage bait again.
Cool, but we already have robots doing this in 2d space (aka self driving cars) that struggle not to kill people. How is adding a third dimension going to help? People are just refusing to accept the fact that machine learning is not intelligence.