halflings

> If somebody creates ClosedAI company with product called ChatLLM, does it mean they can start sending takedown notices to everybody left and right who is using "LLM" in their name?

Those things are nothing alike.

GPT is a very specific family of models, all created by OpenAI. The copy-cats came after OpenAI released those models, specifically to point out similarities with the OpenAI-created models.

LLM is a generic term and cannot be trademarked.

You can trademark "Coca Cola", but you can't trademark "Cola" or "Soda".

> why would I use a clone instead of the real thing

Has nothing to do with why people don't use Edge (they still didn't use Edge when it was using a custom engine; same for IE before it).

And it's only the techy minority that even knows Edge is built on top of Chromium.

> cheaper in the long term

Citation needed :)) There are economies of scale and various optimizations that are just not possible with dedicated machines.

Anecdotal evidence: a company I worked in had a dedicated DC with hundreds/thousands of machines that mostly ran SQL queries on petabytes of data (any query would take ~5-30 minutes). Eye-watering budget and whole teams to maintain the cluster... They switched to GCP/BigQuery, got queries that ran in seconds at a fraction of the budget.

Responsive? I have to scroll past the left column ("Site Navigation") to read anything. That's the opposite of responsive (e.g. 2 columns on wider screens, 1 in mobile).

Examples of such architectures:

RETRO (Retrieval-Enhanced Transformer) https://arxiv.org/abs/2112.04426

And much older (2016): Differential Neural Network https://www.deepmind.com/blog/differentiable-neural-computer...

The scenario you gave is very unlikely for stocks with large trading volume like TSLA. The spread is nowhere near that wide, and if it was market makers would make a profit out of that.

> even if the link goes to Spotify, they do a ton of things to siphon listeners that came for my music away from listening to my music, including NOT playing more of my music after the intended song plays

If always playing songs from the same artists is what people wanted (e.g. lead to more overall listening time), Spotify would 100% do that. You can't pick what a radio station plays next after starting one of your songs, likewise Spotify gets to pick what their users prefer (visibly: not always songs from the exact same artist).

+1, the most insulting people I've met were the young "entrepreneur" types, not older/accomplished/actually wealthy people.

The paper doesn't mention any baseline, and so claims like "86% accuracy" are a bit useless: what if players do the same action 85% of the time? Then the model wouldn't be not predicting much. If on the other hand the most common action happens only 30% of the time, then that would be an incredibly strong result.

This is most likely just a bias in the sample, rather than actual increase in the median page size on the web as a whole.

(increases in median size are highly correlated with the sudden jumps in the # of pages indexed, so they likely started covering different subsets of the web that have a different distribution of page sizes)

Don't take it personally, but you seem to hate quite a lot of things that are outside of your control, or just objective truths that are there to stay:

* DevOps: If people find value in releasing software in an organized manner to avoid breaking websites / "testing in production" and losing a lot of revenue, why do you strongly hate or care at all? You don't have to work on devops.

* Learning new frameworks: Did someone force you to learn every new JS framework released in the past 10 years? React is the main framework people have been using since ~2015-2016, and other frameworks are either not super important (latest CSS framework someone came up with) or worth the cost of learning (Gatsby, Next). Continuous learning is an important skill in all jobs.

* "having the fundamental stuff I learned early on - objects are good, always separate presentation from logic - being totally flipped around": People realized that this is not the right approach for building web frontends. These best practices are not meant to be seen as dogma/ideology, they adapt to people's experience (and usually will only apply in some contexts)

So I think you can just relax, enjoy the parts you love, and be more open to things changing around you.

Most of these games are made by publicly traded companies, which have to release information such as dev/marketing costs to their shareholders.

What do you mean by "UI component driven version of this" exactly? Like this, but to build UI components somehow?

FWIW, there are plenty of graph-based dev tools, especially in the data science & ML world, and they seem to only fit the nerdiest of audiences (like mine!) ; not very practical for most people.

> But there is a huge double-standard here about what a job is and how compensation works. [...] Why is a VP paid $500,000 for taking care of their child for 6 months but an entry level employee is paid $50,000 for taking care of theirs?

You can't look at a VP's output the way you'd look at a factory worker's. The VP is being compensated for the work they'll provide over the time spent at the company. They might have no output for 6 months then provide outsized output during the next 2 years (as organizational changes they've made start paying off).

Their compensation is based on this lifetime output. Them taking parental leave in average once every 4 years is taken into account in this formula.

People that never have children are still a minority, so they get lost in this calculation. Yes, they're definitely being "underpaid" in that sense the same way someone leading a healthy lifestyle is "under-benefitting from his tax contributions".

Same goes for agave syrup and other health fads.

That's on the roadmap! Will give more concrete examples.

You're right, it's way too subtle. I'll fix that!

A bigger issue is I realized the weights downloaded do not include the data preprocessing... So proper model export will unfortunately take more work before it's fully ready to use.

Thanks! Ease of use (e.g. no lengthy signups / waiting minutes for cloud jobs to schedule) is one of the main features, so I'm happy you appreciate that!

RE:bias/explainability, that's indeed the main argument against asking people to use ML as a black-box.

I think fully understanding the very nuanced biases that can sneak in data (e.g. selectivity bias from some events being more represented in records, let's say) does require a keen sense for data, which an automated tool likely cannot provide. My bet is that this is not enough of a reason to block people from at least dipping their toe in the ML world, and that we can do more education down the line (e.g. about evaluating systems in real-life, to at least catch underperformance before looking for its source) to solve these cases.

On explainability, some basic tools (e.g. feature importance) are in the roadmap, I hope to get to it in this quarter!

RE:creating datasets being the hardest part of the job, not modeling... well, I think you're 100% right. And that's a tougher nut to crack.

One thing I'm planning to do to help here is to provide a number of "templates", e.g. concrete use cases that people can piggy back on. e.g. explain to realtors that they can estimate house prices by creating a spreadsheet with features A,B,C and D. I can't do this for every imaginable use case, but I hope this is enough to at least inspire people on how to think about data and ML.

A thousand times yes! :) This is my entire bet with this project: People are gate-keeping ML usage behind "you need to first understand the math and take a course in statistics", yet no one would ask you to "understand how compilers and CPUs work" before writing a simple mobile app.

Knowing the underlying math and engineering is of course still useful for those that want to go the extra step (squeezing out .5% accuracy, or training a model on 1TB of data), but for all normal people out there ML should be as simple as loading up a webpage and dropping your data.

Thanks for the support!

Try to use either of those solutions, and you'll be hit with a "Schedule a call/demo" wall, have to chat with their sales team, etc.

You can train a model in <30s with ML Console. Two clicks: 1 to pick a dataset, 2nd to click on "train a model". No wait time or sign-up required.

There's a very good reason for this: ML Console uses a very different approach, as all computations (data pre-processing, model training, etc.) is written in Javascript and runs in the client.

This makes the app orders of magnitude more responsive than any competitor (and hosting it is so cheap that I'm offering it for free during this beta phase, something these services wouldn't be able to afford).

With that being said, cloud-based solutions like these will always have their place when you need to train on terabytes of data, or squeeze out 1% higher accuracy with some state-of-the-art models.

BigML and AutoML are great! And I myself write a lot of custom TF code.

I see this as addressing a different stage (rapid/cheap prototyping) and audience (less technical people, smaller enterprises); it'll never be as powerful & state-of-the-art as advanced ML tools by cloud providers, and that's OK :)

> I'll give this a try, though I'm not sure what strategy it's using (I'm assuming an ensemble?)

It's embarrassingly simple right now: a fully-connected DNN with a static architecture (hand-picked because it worked fine for all datasets I tried), but you can also enable ensembling and/or change the network's architecture (just click on "Show advanced settings")

> Also you're asking the user if the objective should be classification or regression, I think that should be automatic, it's already buzzwords for your typical marketing person :)

Good point, but it's not always something that can be determined automatically. I did add some heuristics to warn people when we think they are wrong (e.g. running classification on a numeric variable with too many unique values, usually the sign it should be a regression problem instead).

One thing that might help here is that the redesigned flow shows a step-by-step instructions where one step is to pick the task (regression vs classification), and we'll add some text there to help people understand what the difference is. I'll also consider the heuristic approach going forward! (and make sure it doesn't override cases where users already manually set the task)

The project is not open source for now. I might open source certain components in the future, but the priority is to finalize the app first.

These are just very common test datasets in ML, didn't look at sklearn when developing this.

Agreed, although the app is still much easier to use than other ML apps I've tried, it's still a bit too confusing/"magical" for non-technical folks.

As mentioned on the parent comment, this is the #1 priority, and I hope a redesigned flow (e.g. step by step from loading the data, picking the target, then features, and explaining things at each step) will help here. I'll also have a page with concrete use cases, including marketing.

Please reach out (email on the website) if you have some ML use case you'd like to solve with ML Console, happy to help you prepare your data etc.

Thanks for the feedback!

1) The target user is: a freelance marketer, small/medium enterprise without dedicated data scientists, technical people (e.g. engineers) from other fields, or without extensive stats/ml knowledge.

2) RE:lack of hand-holding. This is likely the biggest challenge for this project: showing people how to think about ML, and how to use it to derive value for their business, without going through hours of training or lengthy tutorials.

> - Guidance on what consitutes good data and how to structure it for input

> - Examples of how it might be applied to their use case Most users are stuck at this initial phase (preparing the data).

One thing I'm working now is adding use-case focused guides: short article explaining how a realtor would go about building a model to help them roughly value houses (including data collection). I hope this helps with these two points.

> Help interpreting of whether the MAE / loss is good or bad

There's a couple things I'm working on that might help: 1) Show metric improvement relative to a baseline (e.g. MAE for a model that always predicts the mean). 2) Show both train and test curves. The current curve is only on test data.

> Better understanding of when to use regression vs. classification

> Some sense of how training and validation is done

I'm currently redesigning the UX around a step-by-step flow (for initial users at least), that should give a bit of room to explain things along the way (e.g. what classification/regression means for total beginners).

> Some automatic way to prevent overfitting

Medium-term models there'll be a mode to continuously train models to tune hyperparameters, that should help avoid overfitting. Until then it's mostly handpicked parameters (including regularization), and having tested this on Kaggle challenges it still sometimes beats my hand-written ML code :)

> Model weights

You can already download the model weights (download icon next to the model name) ; or do you mean feature importances? That's a planned feature, but it's not straightforward to implement in a generic way so might take a month or two before it's shipped.

ML is still mostly used by tech companies, but non-technical people are often those that would benefit the most from it, e.g. realtors training small models to get better house estimates, marketing folks optimizing their ad campaigns, Shopify sellers sourcing new products, etc.

Existing no-code ML solutions are quite expensive and way too complex for small companies: long sign-up flow, require you to read 3 tutorials and go through 26 steps, spin up cloud servers, learn a lot of jargon... and the UX is often optimized for technical people, not digital marketers or Shopify seller.

ML Console takes a fresh shot at this problem: 100% client-side ML training. This is enabled by modern web technologies (WASM, WebGL), which allow us to process data and train models with a minor performance overhead.

AFAIK, this is the fastest way to train an ML model compared to all other solutions (cloud-based or not):

1) No lengthy boilerplate code for every new project. 2) No downloads. 3) No sign-ups or credit-card checks. 4) No lengthy tutorials. 5) No need to spin-up cloud instances and share data back-and-forth between the client and the backend.

Best of all, this means we never see your data, as all computations run locally. This also allows us to be cheaper than all our competitors => for free :) we'll provide a subscription service offering more advanced features later on.

We're still very early stage, so any feedback would be greatly appreciated!

Not for now, might open source parts of it as the project progresses!

Just like people don't have to learn Python and Pandas to build a simple spreadsheet, they shouldn't have to wrestle with complex libraries / cloud tools to train a simple ML model.

This app allows you to drag-and-drop a CSV, pick columns to use as inputs or the target, click a button and boom, a model is trained. More advanced users can customize modeling details (optimizer, hyperparameters, etc.).

This might surprise some, but this 100% running in the browser, with no backend whatsoever. Even the machine learning bits are in Javascript (using tensorflow.js + other libs).

For JS skeptics, yes there's a performance cost vs cloud-based services, but try to use any of the ML tools advertised as "beginner friendly", and you'll realize that this cost (~30s) is much lower than the time sunk into setting-up a cloud account, adding billing info, reading lengthy tutorials... just to train a linear regression model.

This also makes it:

1) 100% private: Your data never leaves your computer, everything runs locally.

2) Free: Hosting bills are cheap, so I will keep offering this service for free, and will only charge for more advanced features.

3) Fast & available offline: No need to wait for cloud jobs to schedule, and it all runs in your browser so you don't even need to be connected to the internet.

Feedback appreciated!

contact@mlconsole.com

This is still very much WIP, and I am now working on a UI to export models and use them on new data.

Nothing fundamentally new, but it's particularly painful in this documentary (or "docu-fiction") though. Couldn't watch more than 10 minutes due to how biased it was, but friends were raving about how great this story was...

> Maybe that's why Google doesn't even list my app even though it's one of the longest running "web apps" alive

Is your hypothesis that Google would have some incentive that you collect user interaction data? Why would that be?

> used to be listed #1 when it first came out

Back in 2002. A lot has happened since then. Your app [1] is not comparable to what competitors like Zoho offer. There's plenty of potential reasons for not being ranked higher: from the landing page, to the increase in SaaS competition, to not keeping up with UX trends etc.

[1] https://ezinvoice.com/