Hacker News new | past | comments | ask | show | jobs | submit | UncleEntity's comments login

> AGPL-3.0, free for research and educational purpose...

...or any other purpose allowable under the AGPL like, wait for it, commercial purposes.


You're right, UncleEntity, thanks for highlighting that. My phrasing could have been clearer. AGPL does allow various uses, including commercial, provided its terms are met.

Our intention with LightlyTrain (AGPL/Commercial license option) is to offer a streamlined, production-ready pretraining engine. This contrasts with our other library, LightlySSL (github.com/lightly-ai/lightly), which is MIT-licensed and geared towards researchers needing flexible building blocks.

We found many companies wanted a simpler "it just works" solution for pretraining, which is why LightlyTrain exists with its specific licensing options tailored for commercial teams alongside the AGPL.

Thanks again for the clarification!


> ...it’s extremely unlikely that Europe will be successful at this. The future is US-China dominated.

I think the events of the last few weeks have been a wake-up call to many people around the globe who now see the value of not having their entire economy (or telecommunications &etc) be vulnerable to the whims of one person, be that the richest person in the world or a president nobody in Europe voted for.

I suspect there's a lot of really smart people who were previously happily living/working in the US who are now looking to emigrate to other lands, Europe could easily allow them to go over there and practice their skills on a Europe First initiative. Could also rope in the Chinese to help as they also aren't too happy with the current state of affairs.


Sure I’m just saying Europe is not a place. Europe is not a giant coordinated economy. I’m “European” and I know how hard coordination problems are in the continent, since I’ve lived through its mismanagement my entire adult life.

Britain tried voting themselves out from under the influence of a politician in Brussels that nobody in Britain voted for, but found that removing that influence was a lot easier said then done.

> or a president nobody in Europe voted for.

Are you talking about the European Commission, which we didn’t vote for?


Your head of government appointed a member of that commission, though. For better or worse, it's similar to that same head of government being appointed by the people you actually voted for in a parliamentary system (but two steps removed).

> the value of not having their entire economy (or telecommunications &etc) be vulnerable to the whims of one person, be that the richest person in the world or a president nobody in Europe voted for.

.

> Could also rope in the Chinese to help

This is pretty funny. That's just switching who you rely on.

> Europe could easily allow them to go over there and practice their skills on a Europe First initiative

In Europe we prefer unskilled labour from third world countries so that they can do uber eats for cheap


> Get Rust performance without having to learn Rust

I do that by using the Python C-API, no rust, no fuss...

Thinking about it, I haven't tried to see how well the robots do with wrapping C libraries -- though I usually use pybindgen to generate the initial py-module then fiddle with the code manually.

Other than that lisp implementation has anyone used the python api which lets you register a module as the 'compiler' for external file types? I forget what it's actually name is.


Yup but you're skilled enough to write--and more importantly, maintain--the required C/C++ code. Most devs and companies we talk to just want to make something people want; they don't care for the added complexity of writing and maintaining native code.

The way I like to think about this is how much more code got written when "high level" languages like C came onto the scene, at a time when Assembly was the default. Writing Python is way (way way) easier and faster than any of the lower-level languages--no pointers, no borrow checker!


> but not worth it if you have other better job options

Pretty much any service job, really...

When I had occasion to take a ride share in Phoenix I'd interrogate the driver about how much they were getting paid because I drove cabs for years and knew how much I would have gotten paid for the same trip.

Let's just say they were getting paid significantly less than I used to for the same work. If you calculated in the expenses of maintaining a car vs. leasing a cab I expect the difference is even greater.

There were a few times where I had just enough money to take public transportation down to get a cab and then snag a couple cash calls to be able to put gas in the car and eat. Then I could start working on paying off the lease and go home at the end of the day with some cash in my pocket -- there were times (not counting when the Super Bowl was in town) where I made my rent in a single day.


Seems like the difference between a profitable investment and a profitable company.

They invested tens of billions of dollars in destroying the competition to be able to recently gain a return on that investment. One could either write off that previous spending or calculate it into the totality of "Uber". I don't know how Silicon Valley economics works but, presumably, a lot of that previous spending is now in the form of debt which must be serviced out of the current profits. Not that I'm stating that taking on debt is wrong or anything.


To the extent that their past spending was debt, interest on that debt that should already be accounted for in calculating their net income.

But the way it usually works for Silicon Valley companies and other startups is that instead of taking on debt they raise money through selling equity. This is money that doesn't have to be paid back, but it means investors own a large portion of this now-profitable company.


> My wife, the farthest you can get from the HN crowd...

She is literally married into the HN crowd.

I think the real AI breakthrough is how to monetize the high usage users.


I have to say, as an American citizen, that maybe, just maybe, we can give just a little benefit of doubt to a scientists who followed all the rules to come to this country and maybe, just maybe, isn't all that familiar with the subtle nuances of the anti-frog smuggling statute.

Back in the day they would just be, like, "yeah, sorry, we can't let the frogs in."


I heard a rumor that China has been dumping some of their dollar-denominated assets as a sort of "bad dog" reaction to the current state of affairs.

Back when I was in the army the objective when doing individual land nav was: go hide in the woods until around the time to go back. Of course there was a bit of walking because we never started and ended at the same place so you had to have some familiarity with finding your way in the woods. Once in a while they would go out and change the point numbers around, usually when it was an actual test, so you had to know actual land navigation skills if you didn't want remedial training on the weekends.

The last land nav course I ever did was in the reserves and in the mountains. Instead of traipsing through the underbrush I just followed the fire breaks and occasionally took azimuths off surrounding mountain peaks and only missed one point because I didn't feel like hiking up the steep hill it was on top of.

Not too sure I see the value in online land nav but I also learned it before GPS was really a thing.


I think the premise of the argument is you don't need to compete in the marketplace because you can just tell the robots what you want and they will happily produce it.

The fallacy that people are making is they look at the current state of things as 'the way it'll always be' while the AI companies, and I mean all the AI companies, are looking to corner the market so have no moral issues with taking out a whole swath of skilled jobs.

The cost to compete with another business's software is a high-end GPU.


Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: