Ask HN: My company is lying about product. What do I do? | Hacker News Reader

Ask HN: My company is lying about product. What do I do?

41 pointsnotsmart46 comments

My company sells product publicly claiming it to use AI. The product contains no AI at all. The statements they make about the product and how it uses AI are 100% false.

What should I do?

View on HN

I had a similar worry about a year ago. On reflection I realised that AI is a much broader, vaguer term than I'd thought. It can be used to describe any software which replaces some human process involving decision-making. A huge range of software can legitimately or tenuously be made to fit that criterion. Doesn't have to involve ML, deep learning, neural nets or anything like that, even if some people would prefer AI to specifically mean those things.

I have gradually come around to the idea that AI as a term is more a statement about the user experience than the underlying technology. If a product "feels" intelligent to the user, and that is not being provided by manual human labor, it doesn't matter whether the underlying system is a lookup table, and rules based expert system, a numerical optimization technique, or a neural network. It is entirely reasonable to call game AI by that term even if the most sophisticated algorithm it uses is A*.

As an engineer, I try to use more specific terms for talking about implementation decisions like "machine learning", "deep networks" or "dynamic optimization".

It might be reasonable to raise a point like "I'm not sure that our current product actually seems intelligent to our users, what can we do to better live up to the marketing vision", but objecting to the term AI just because it means something different in engineering than in plain English seems a little bit unreasonable. If the target of your marketing is engineers, that might change things.

Remember when "expert system" was the AI buzzword du jour?

Technically, the majority of expert systems were narrowly focused (single purpose) search engines that were unable to extract or infer novel information from the collected data. That required humans who still referred to it as artificial intelligence.

I was confused when I learned linear regression is considered ML now.

Its been considered ML ever since experts started writing books about ML (and before that, "pattern recognition"). Most introductory ML books cover it in some detail.

Well, it's not like a support vector machine that just separates two point clouds by a hyperplane is that much more advanced than an algorithm that draws a line that best matches a point cloud.

Let’s get one thing straight: If your company is saying they use AI to solve their customer’s problems but they do not, that is lying. 100%. It’s fraud. You are absolutely right to be concerned.

I would argue that you have a moral obligation to do something about this while bearing in mind your own needs and obligations (e.g., as a breadwinner for your family).

Others have been in your situation before. Try seeking advice from Tyler Schultz, one of the first Theranos whistleblowers: https://www.linkedin.com/in/tyler-shultz-450923126

There may be others, I’d love to see comments giving other suggestions.

You could also get in touch with the Wall Street Journal— tip line is here https://www.wsj.com/tips

You could send a message directly to reporter John Carreyrou, the WSJ reporter who worked on the first big Theranos story.

Going directly to the press has some serious risks — legal risks especially. That’s why you might first begin by merely seeking advice and input from people who have worked on this before.

Many people will tell you that the legal risks mean you shouldn’t do something. I understand that position. At the same time, I believe we human beings have an obligation to each other to deal fairly and honestly, and to help bend the world in that direction. It’s not an easy position for you to be in. Good luck.

This is silly. Companies call all sorts of things AI that aren’t - from IBM to your smallest startup. It’s a nothing phrase

furthermore, it's not a lie if they believe it

It's possible OP is in a Theranos type situation, and that's a good thing to look out for.

It's also conceivable the company is doing Wizard of Oz exercises (e.g., having humans simulate what they ultimately want software to do), to refine their model and build ML training data, which is an entirely valid approach, but they'd have to be upfront about that.

Upfront to users, and especially, investors. For example, if someone were telling an investor that they have a DNN servicing the customer requests, but it's actually humans, that would create a very difficult situation, and fixing it would probably include consulting a lawyer. In that scenario, there might be others in the company who are pissed at the person who ran off their mouth, but not want to take a hit of fixing it in a good way (and then things can get worse; "the coverup is worse than the crime" is a thing).

I suspect OP's situation is much more easily corrected, or might even just be a small misunderstanding.

Wizard of Oz experiments don't work if you tell the user (from a user study perspective). You could tell them afterwards, but depending on the product / service that may not make sense.

This isnt necessarily an experiment, users' perception of how might not be a concern at this time, and there's different kinds of Wizard of Oz exercises.

For example, when I first learned the term (in HCI, or human factors engineering), they used the example of a mockup by photocopier designers, in which obviously the people using it knew there was a human moving sheets of paper to the slots.

Maybe there's a better term?

Try to see how hard it would be to make it true, even if it's for some tangential accessory use. Add some elementary AI function to add a line to a log in your build scheme if that's something you have control over.

Then, if you jump ship, you can put "conceived of and implemented 100% of AI capabilities in former employer's operations" in your resume and cover letter. Win-win.

You are not alone

https://www.theverge.com/2019/3/5/18251326/ai-startups-europ...

> Forty percent of ‘AI startups’ in Europe don’t actually use AI, claims report

I'm pretty sure that it's not only related to Europe ;)

In the modern market, "if statements" and "control flow" are "AI" in the eyes of marketing execs. It probably won't do you any good to try and tell them otherwise.

In games this is widely known and accepted even, but I guess that's a slightly different case since there's some history behind the why. It's AI because the actors are game characters trying to emulate a real behaviour, whether that's a soldier, dinosaur, or a fish. "Artificial intelligence" makes sense.

I don't think it's not too bad to call a complex decision process "AI". ML is a separate distinction.

If you can't take the lies, leave. It's the only way. I would refrain from making this public because it could get you in a lot of trouble, which isn't worth it.

What sort of trouble beyond getting fired?

No idea about the OP, but in my case every contract I signed contained some clauses that prevented me from speaking or acting in any way against the companies interests, at least while I was working there; just like a sort of non-compete agreement but applied more broadly. Luckily I never had to test those clauses, but I'm sure had I done anything against them they would have destroyed me in court in no time.

If I was in the OP shoes I'd leave ASAP. One day someone will let the cat out of the bag anyway (coworkers, researchers, competitors doing reverse engineering, etc) and he could lose his job anyway if the company tanks. Not to mention being related to a technical fraud.

Most likely the way he knows it doesn't use AI would be covered under the NDA that they signed.

Then they'd get sued for the potential revenue loss and damage to company image.

He might win, however corporations have much deeper pockets for lawyers than your average citizen.

You need to share more context.

TBH, a lot of AI does not do things better than a human. It merely does them more cheaply or more scalably. Thus if you are prototyping or building an mvp, it's not necessarily terrible to have your "AI" be a human. Even the best AI systems are often a mix of humans and software, where eg the humans label or the ML code sends examples near the decision boundary to humans for review.

Building an initial system that is human operated and transitioning that to machines is how a lot of companies are built.

Of course, if you're lying to your investors, that will not end well. Ditto future employees.

This is tough. The lie is probably originating from the executives in your company, so you probably can't just nip the issue in the bud and change the verbiage of what you do. So, with that being said, you probably will have to start by voicing your concern to your most immediate superior. When you do this, don't immediately jump in as a whistle blower who is trying to implement justice (which I am sure you were not going to do, but just in case).

Rather, have an honest conversation, and see if there is anything that can be done to create a more honest image of what you are doing. Hopefully they will have some sway or insights on how change can be brought about.

If your boss is unwilling to hear you out, you might have a problem you can't fix. In this case, if the company is willing to deceive its customers for financial gain, then it is likely time to start looking for a new job elsewhere.

In the long run, companies that do shady business practices get found out, and won't last very long. Luckily for you, it seems like you have an honest head and your shoulders, and will be a great candidate for a business that does things the right way.

Personally, I wouldn't worry about it, my rice cooker has AI. If it really bothers you personally, leave. I wouldn't consider it a moral obligation to expose the company unless there's more than the usual marketing and biz aspects to it, e.g. individuals being put at risk or exploited. I don't include investors who's job it should be to sniff this out.

Keep in mind a simple if statement in code is AI. AI has always been a (shitty) marketing term, and sales and marketing has been trying to use it to sell products for the last 50 years. You’d need to elaborate a lot more on the specific claims of the company before I’d personally be worried about sales and marketing selling snake oil with this term...

Look for another job in the background if you want and take it after you found a better one.

Also what's more important is whether the product is useful or not.

At the same time it will be hard to find a company in which the CEO doesn't ever lie :(

I suspect many of the AI companies floating around nowadays are really just faking it till they make it on the assumption that they will figure it out once they get big enough.

In short: GTFO. Longer: They're unethical and eventually it will be discovered and it will tarnish the work in your career. Do not stay and be their accomplice.

The team is small enough that you should be able to talk to whoever is in charge of either engineering or marketing and ask them about the strategy and positioning. In some cases you need lots of data to start making things work with and kind of ML; this could be that period of time.

If it violates your integrity to work there after you have an answer then just move on.

Any time I read or hear "AI" these days I automatically translate it in my head to the word "software". After all, all software is a kind of artificial intelligence, so it's not exactly dishonest to use the term AI -- it's just a bit lame.

The benefit of doing this mental translation is that I'm no longer annoyed all the time ;-)

Since you are not fine with it, you have two options:

1. You quit right away. 2. You force yourself go to work and eventually burn out then quit.

Unless you are in a position that can influence the company, it doesn't worth your time to change it.

If you need the job now, invest more time and resource on yourself so you can move on to other place easier later.

Is it an aspiration for the company and are they close? Is it miscommunication between product and marketing? Or is it intentionally lies? Are you offering some sort of information automation? To me I can see anyone not in computer science easily being confused. So maybe intention here matters

I have yet to work at a tech company that doesn't either maximally misrepresent or outright lie about its product offerings. Incidentally this also seems to be a good way to identify companies that are likely to fail or underperform in the long run.

who's being fooled ?

- the investors ? most likely they know your four man team is not capable of doing any solid AI work and are aware that your CEO is bullshi$$ing but they are actually betting on this ability to sell what does not exist being a personality quality.

- the customers ? why should they care. if they care, they should ask where's the ai. "show me the ai"

- you ? your not being fooled. you might have been when being signed up for the job, but now just accept it. if AI becomes a thing it might eventually find a way into your product. otherwise id just play along now while the cheques keep coming.

I may not agree with these viewpoints (I'm not the OP so I don't have to form an opinion) but they are unique and worth considering.

Any chance you're working for a legal "AI" platform?

Nothing. Take your salary and do your job, assuming you aren't the one stating the lies.

Being a martyr is cool and all, but I'd rather continue living.

Do you think they'll kill him?

Of course not, but getting sued into oblivion might as well be a death sentence. Why take the risk, when the potential reward is so small?

Show me a business that always 100% tells the truth about their products? Anybody?

Sounds like my perception about majority of vendors at recent RSA conference.

Is anybody concievably being hurt by the claim that it's AI?

AI is just a marketing buzzword. Don't worry about it.

Does it at least solve the problem that it says it solves ?

Do you work for IBM, by any chance?

leave.

Reply on news.ycombinator.com