If you are looking for a open source Chat WebUI with support for OIDC, maybe you are interested in the one we are building?[0]
We are leveraging oauth2-proxy for the login here, so it should support all OIDC-compliant IDPs, and there are some guides by oauth2-proxy on how to configure for all the bigger providers. We do have customers using it with e.g. Azure, Keycloak, Google Directory.
I see you have a dockerfile.combined - is this built and served via gh artifacts? I can try it out.
Pros: Open source, and focus on lightweight. This is good.
Cons: "customers" - Ugh, no offense, but smells of going down the same path as "open" webui, with the services expanding to fill enterprise use cases, and simplicity lost.
LLMs.py seems to be focussing purely on simplicity + OK with rewriting for it. this + 3bsd is solid ethos. Will await their story on multi-user, hosted app. They have most of the things sorted anyway, including RAG, extensions, etc.
> I see you have a dockerfile.combined - is this built and served via gh artifacts? I can try it out.
Our recommended way of deploying is via Helm[0] with latest version listed here[1].
> with the services expanding to fill enterprise use cases, and simplicity lost.
TBH, I don't think that simplicity was lost for OpenWebUI because of trying to fill enterprise needs. Their product has felt like a mess of too many cooks and no consistent product vision from the start. That's also where part of our origin story comes from: We started out as freelancers in the space and got inquiries to setup up a Chat UI for different companies, but didn't deem OpenWebUI and the other typical tools fit for the job, and too much of a mess internally to fork.
We are small team (no VC funding), our customers end-users are usually on the low-end of AI literacy and there is about ~1 DevOps/sysadmin at the company our tool is deployed, so we have many factors pushing us towards simplicity. Our main avenue of monetization is also via SLAs, so a simple product for which we can more easily have test coverage and feel comfortable about the stability is also in our best interest here.
Honestly - the video barely touches on this at all, despite making it the "hook".
I was pretty disappointed that he doesn't discuss the EMAS mechanics, structure, actual stopping distance, or impact to the plane in nearly any real way.
He does show a LOT of animation of layered runways, which are mostly not that informative.
There is some decent discussion around maintenance and material choice, and some very basic discussion of infrastructure requirements outside of the runways themselves that's... ok.
Overall... I thought this was a solid C+ video. It shows planes plowing into an EMAS, then does jack all to discuss that, while bringing up a lot of less interesting discussion of runway building (which despite the claims in the video, do actually correspond very highly to how we build highways, just with different weight/maintenance requirements.)
I don’t think anything you said is wrong, but I do think you’re misreading the intent of the channel.
Practical Engineering is very deliberately framed as edutainment. The animations, pacing, and level of depth are conscious choices meant to keep non-specialists engaged rather than to maximize technical rigor.
In that sense, it belongs alongside other popular "science-y" channels like Kurzgesagt, Vsauce, and Technology Connections: content that prioritizes narrative and engagement over completeness or instruction.
The target audience is the broad middle of the technical bell curve. Animations of runway layers may make the video less appealing to you, but more accessible to a much larger, less technical audience.
Different goals imply different success criteria. If the goal is reach rather than comprehensive education, a million views in seven days looks like success.
> I’m not sure how reasonable it is to expect a business to continue offering the same product or service indefinitely if market forces are pushing it elsewhere
Market forces aren't pushing it elsewhere. The cornerstone of Microsoft still is Windows and Office. If those would not exist nobody in their right mind would choose Azure over AWS or GCP.
By letting their guard down on those fronts and letting Windows and Office degrade more and more, they are exposing themselves to the risk that someone ends up building a competitive company filling those niches and people risk the switching cost in order to get away from ever increasing Office 365 subscription costs.
Agreed. Companies pick Azure because they have already invested in Windows and Office. I have never worked in one company that uses Azure but not Office. They usually buy azure because of the discount.
> nobody in their right mind would choose Azure over AWS or GCP.
There's a really interesting dynamic here in that Azure has a solid spoiler role for large organisations that don't want to be commercially dependent on only AWS, and they can probably get really solid discounts if they're aready on board elsewhere. It's something that doesn't play out with Microsoft's other products nearly so much: you get shouted down if you want to have desktop diversity, but having a multicloud strategy is (in my experience) looked on as essential.
The cornerstone of Microsoft still is Windows and Office.
Again, is it really, though? I have no special insider knowledge so perhaps this is just a misunderstanding of the public information, but just going by the organisation structure, leadership comments and recent financials, it looks like Windows makes up a relatively small part of Microsoft’s revenues these days, while the traditional desktop Office applications seem to be almost lost in the noise. The emphasis seems to be firmly on cloud services, though admittedly with all the rebranding from Microsoft lately, I find it hard to understand even what basic products and services they offer any more.
My point is that Windows/Office are a essential part in their sales funnel.
Google also makes most of their money in "ads" but if they were to axe Search and Youtube (which in an reduced view are only sales funnels for ads), they wouldn't have much of a business left.
I expect you’re right about the sales funnel angle, though neither Windows nor Office seems to be the same kind of product that those brands have traditionally described any more, presumably for that same reason.
Windows appears to be positioned more as a platform to reach all the online services now, rather than its traditional role as a desktop OS. Can you even activate it without being online and having a Microsoft account any more? I’m out of the loop, so genuinely don’t know the answer to this one.
Office — or whatever it’s being called after the recent changes — also appears to have morphed into something quite different. I tried searching just now to see if you could still buy a permanent licence and install the classic applications like Word and Excel locally, and some sources implied you could, but I didn’t actually find any way to buy it in five minutes of looking around office.microsoft.com. As far as I saw, that site is now 100% about the online SaaS version and trying to get users to save their documents in the cloud. For businesses, the strategy seems to include promoting other online services like SharePoint and Teams as well.
So I think I stand by my original argument, though I don’t think it necessarily disagrees with yours. Windows and The Software Product/Service Formerly Known As Office might still be a significant part of Microsoft’s sales funnel, but they aren’t the products that Windows and Office used to be any more. The products they used to be have been repurposed to support an online-first corporate strategy, along with almost everything else in the Nadella era. Would Microsoft care if 100% of their customers stopped using Windows tomorrow and jumped to Apple or Linux systems, as long as they still used the other services that generate most of Microsoft’s revenues these days? I’m not entirely sure they would.
Not much? Google would still have cloud services (which unlike Azure adoption depending on Windows/Office, only basically depend on the Internet), Gemini and Google Drive paid subscriptions, their flagship Pixel line...
They kinda did the same with IE. Got into a virtual monopoly and then let what was already a mediocre browser crumble into a pile of manure. So yeah then Google came along and ate their lunch. They should know better at this point.
I think similar to Joe Rogan that's the main value he provides to listeners. He identifies guests that have some veil of intellectualism and provides them with a platform to speak.
However I don't think that makes for an interesting interviewer. There are no challenging questions, only ones he knows will fit into the narrative of what the guest wants to say. I might as well read a 2-3 hour PR piece issued by the guests.
What you call "platforming" I often call "listening to what someone says/thinks". Not every interview needs challenging questions, or to be a battle/debate, and sometimes it's not appropriate (above George Hotz being an example, difference in qualifications being another). But, I enjoy trying to understand someone, quirks and all, especially the human aspect, flaws and all. It's interesting seeing the differences in people.
From what I've seen, people that crave "challenging questions" usually most enjoy activist interviewers that are very strongly aligned with their own (usually political) worldview. I don't think that describes Lex Fridman, or me as a listener, at all, and that's fine.
No, not every interview. But if an interviewee presents fiction/hatred as fact the interviewer should have the ability to call that out or at least caution the reader with a "I don't know about that".
A specific example that comes to mind is Eric Weinstein's appearance on the podcast and letting him talk about his "long mouse telomere experiment flaws" without questions which at that point had been thoroughly debunked.
I find little interesting "human aspect" to be found therein, as it usually boils down to "you are lying (to us/yourself) for your own gain", which isn't novel.
There are podcasts that do a similar long form format well. A great example is the German format "Alles gesagt?" (~="Nothing left unsaid?"), where interesting personalities can talk for however long hey want, but the interviewers ask interesting/dynamic follow up questions, and also have the journalistic acumen/integrity to push back on certain topics (without souring the mood).
> letting him talk about his "long mouse telomere experiment flaws" without questions
This requires that the interviewer is as knowledgable as the interviewee (the qualification problem I mentioned). Unless the questions and answers are known ahead of time, it won't be possible to know everything an interviewee will say. Assuming this is the case, how should he have handled that response? Should he not interview people outside of his own expertise? I think one way would be "is there any disagreement?" but then you're left with the same problem.
I think Lex Fridman not knowing much about the history/current state of rat telomere research is entirely reasonable. I think a requirement of knowing the entire context of a person is not reasonable. I also don't think it's reasonable to believe everything you hear in an interview, from either human. "Charitable interoperation, but verify" is a good way to take in information.
Sadly their whole frontend seems to be built without QC and mostly blindly assuming a happy path.
For the claude.ai UI, I've never had a single deep research properly transition (and I've done probably 50 or so) to its finished state. I just know to refresh the page after ~10mins to make the report show up.
I've heard this argument a lot, but is the bus driver really more than a rounding error on the balance sheet of the transportation company? I have no experience in this field, but I would imagine the infrastructure for buying and maintaining a fleet of busses, creating a route network, dealing with ticketing, dealing with disruptions, etc, etc, makes up most of the costs around running a public transport company?
I think the benefit to transportation companies would not just be about salaries, but also about predictability in staffing. I've experienced a few times in the past that a strong flu season causes specific bus lines to be serviced with reduced frequency due to staffing shortages.
But concretely, regarding the staffing costs, if I roughly read the financial report section correctly, it looks like for the Berlin transport agency, salaries make up half of all expenses[0]!
Given that that's such a big portion, I think autonomous buses could likely unlock a lot of mobility in big cities by having more flexibility in creation of additional bus routes (as you don't have to consolidate multiple routes into one because you have to pay drivers for each).
It reads like an anti-ad for both. "I didn't use the Copilot IDE because I lack control over the context provided" and "I used Copilot 365 because it for sure doesn't have any context of anything because connecting things to it is hard/expensive".
With experience and optimal prerequisites (good connections to a notary, single founder with default bylaws and no asset transfer into the company) you can do it in ~4 days, e.g. for a holding company.
I did it in ~2 weeks last year, where almost a week was caused by the coworking space I rent at not notifying me of the physical mail from the court. If that physical mail would be eliminated from the process you could probably do it in 2 days.
Apart from that, for any non-trivial situation, the majority of the time will be determined by how fast you can proceed through the process of adjusting your bylaws, etc. and evaluating tax situation (so lawyer + tax advisor waiting time).
(after that the process of waiting for a tax ID starts, which depending on where you live can easily be the slowest part and take ~6 weeks on its own.)
> Seems like such a low-hanging fruit, especially given the very visible success of companies like Framework.
Is there very visible success of Framework? How many people in your everyday live have you encountered with a Framework laptop?
I love there mission, but Framework from all the feedback from users online seems to still be a product that you'll only buy if you put sustainability over performance/convenience.
> a YC-backed attempt
If any successful attempt would be launched, there would be no reason for it to go through YC. In the mass consumer hardware market their little funding and the network they provide doesn't do much. I would strongly assume that a challenger would appear in a similar form as it did with framework with nrp.
> Is there very visible success of Framework? How many people in your everyday live have you encountered with a Framework laptop?
A company that captures most/the entire market is not what is being asked for. Only a financially viable company that provides value to people looking for a certain type of product. Framework is certainly that in its own domain. And something similar could be built for phones.
At what price though? There are many people who say they would buy a phone like Fairphone, but not at that price, or not unless it had a 3.5mm headphone jack or a better camera, etc etc. Talk is cheap but sustainable phones are not.
So are you buying a Fairphone right now? Because from my rough estimation Framework and Fairphone are about the same when it comes to performance/convenience tradeoff right now.
I mostly focused on the "YC disruption" part in GPs question without considering whether there is actually an opening for a disruptor. I think Fairphone may already be filling that gap.
I don't see a reason to buy a new phones when I can buy refurbished flagships from a couple of years prior. The fact that this is the best option available is precisely what strikes me as being so straightforward to fix.
So what would your straightforward "fix" look like? Buying refurbished phones is nothing to "fix", and undercutting the psychological effect of pre-ownership price-drops is essentially impossible.
>So what would your straightforward "fix" look like?
A repairable powerful phone with replaceable battery and pathways to minor upgradability that you can trust not to fuck you over long-term. I.E. the closest equivalent of Framework to smartphones.
We are leveraging oauth2-proxy for the login here, so it should support all OIDC-compliant IDPs, and there are some guides by oauth2-proxy on how to configure for all the bigger providers. We do have customers using it with e.g. Azure, Keycloak, Google Directory.
[0]: https://erato.chat
reply