Though most Wyze cams are now IP65 rated against water and dust intrusion, mounting them upside down outdoors is not recommended (unless sheltered under an eave or similar.)
Water often intrudes and permanently fogs or stains the lens.
Another âdonât doâ would be that thing with Cam Outdoor where you brick it if something something.
I like the addition of the 
. 
To reduce a working electronic device to an inert lump. Unrecoverable.
Often used precipitantly to describe a device that seems grievously wounded. Recovery may still be possible.
AAAAAGH! This firmware update totally bricked my cam! Damn the Wyze! Damn!!
The spirit has left the camera.
So if sideloading is what it is, is the regular way of doing it frontloading?
What do you mean by the future? They are already the present.
More than half of B2B companies have automated chat bots implemented. B2C companies are near 50% already.
Some business industries are already up to near 3/4 adoption.
Overall, when more than half of medium to large companies are already using them, Iâd say itâs no longer about the future, but the present.
Some reports estimate that sometime during 2025, the adoption percentage in general will be as high as 80%.
The future is already here my froggy friend.
The main question is how much it will improve. Also, not all implementations are equal. I was recruited to help one very large organization (they have double or triple triple digit billion$ in just Treasury and market equity holdings, not counting other assets) with their chatbot. I was asked to try to break it or exploit loopholes so they could address them and keep it on target for their intended use case to stay on topic and within a limited scope. I can tell you that some companies have put a whole ton more effort into implementation of their chatbots with a ton of refinement than others have. Some are much more helpful than others. Some even have some agentic abilities and properties, including some having access to different API functions. Others are nearly useless generalists without a useful knowledge, base included as their primary library.
I have some other clients that are in the process of building new startups that have a core foundation built on generative AI. Some of them are actually really interesting That a lot of anti-ai people would approve of as a positive step. For example, one company Iâm helping out, one of the main reasons they put the generative AI as a foundation, to then build a sort of firewall that locks out the AI from ever being able to access any personal information at all. For example, if it needs to review anything that would have a name or phone number or address or payment information or anything else that people would consider sensitive information related to them, those fields are hashed into something totally unrecognizable before it is given to the AI to review, then the AI can do its analysis for things and send it back to the system, which then decodes the hash back into the original information for the human to be able to understand it. There is a complete blockage of the AI being able to access any personal information in any way. Is complete. Separation should honestly be totally required for pretty much every company but especially industries like healthcare etc. right now everybody just lets the major generative AI companies get access to everybodyâs data. And in some cases they will sign a baa agreement or something, but the problem is that there have been a lot of accidental leaks. Sometimes the AI has shared confidential stuff with random other people. So the idea here is to never allow the AI tab certain confidential information in the first place. Then, in addition to hashing critical information for various kinds of data analysis, local model hosting should be something very important for most of the businesses in my opinion. If companies would do a lot more of both of those things, I think a lot of people would feel a lot more comfortable AI having access because data analyzed by a local hosted model canât possibly go anywhere else, and the extra safeguard is that critical data, especially any identifying data is never available to the model in the first place.
After that core foundation which I think is critical, then companies should have their chat bots. Follow various kinds of looping instructions with mild to moderate agentic permissions to interact with the API. If companies did all of the above, then I think in some cases I might even prefer a chat bot over a human for a majority of my interactions. Interactions assuming the chatbot has the authority and ability to process things that I need such as refunds, corrections to services, replacements, warranty processes, etc. if there is something that has a clear outlined policy that any human would have to follow anyway, I donât have a problem with a bot being told to carry out the same policy. A human that reads scripts and a human that follows a set policy is basically a human bot to me already anyway. But at least with the the digital chatbot, it will be consistent and predictable and do things right more often. It should also have a policy for when to escalate to a human supervisor in rare cases. I really donât have a problem with any of that. Quite the contrary.
Exclusive. Oneâs only option for support. 
Interesting. Thanks for expanding the subject. 
Moving as fast as we will be moving, trust placed âreasonablyâ in a company one has experience with (like Wyze) may no longer serve to protect the customer. Competition will demand the company make concessions to the market and people will relent as resistance becomes increasingly inconvenient and impractical.
The trend may not be our friend. 
I donât want to veer too far off topic, but this seems like a big chunk of the problem with Wyze Support, at least in my interactions with the human agents. I donât know how many times Iâve received a reply that doesnât actually address my direct questions and strongly indicates that the human responding with the script didnât comprehend or read at all what I had previously written. Theyâre not using the 
, which is what I expect a human to do. A properly âtrainedâ AI system might actually have an advantage here, because in theory it should actually parse what a customer has written and generate a response that incorporates the context. That might actually be a positive move if itâs done well enough, and then hopefully the actual humans would be given the freedom to exercise critical thought in solving other problems rather than just regurgitating irrelevant scripts.
I think it is.
However, I can also agree that there are both really good and appropriate uses for it, and inappropriate and negative or even abusive uses for it.
But the same can be said about almost anything major. You could say the same thing about money. You could say the same thing about the internet. You could even say the same thing about water, which even makes up 60% of our existence. Cars, fire, electricity, medicine, nuclear power, even language itself. Knives, genetic engineering, robotics, social media.
This is a classic example of whatâs called the dual use nature concept.
I believe it is progress. Even though many of the things listed above also have strong negatives and potential for inappropriate or abusive uses, the accountability for that falls on the person who uses it. Abusively rather than the thing itself which is neutral.
In the end, if chatbots end up being that big of a negative, then the market will decide that and force it to change. But I think that opinion will overall be a fringe outlier rather than the standard opinion. People can absolutely go to a competitor that charges more money to do things differently, but I think the overwhelming market share will determine that they value efficiency, consistency, unmatched availability and instantaneous response, elimination of wait times, handling high volume repetitive tasks (save humans to only have to deal with the complex, high value, or emotionally charged situations), infinite scalability.
Having worked in a call center through college, I can tell you that there were policies that incentivized some employees to do things like cramming and slamming peopleâs accounts so they could make commissions or not get fired by not meeting goals, doing things the customer didnât ask for or purposely worded them in a way to have miscommunication, or in some cases there were people who are stealing payment information from customers (such as when the The customer reads payment information to them over the phone to in the system, some of them may save it personally to use later, I knew of people who got in trouble for things like that), or other weird stalking activity and stuff. Or if a customer sets or offends customer service representative , there are a lot all over the Internet who share about all the ways they sabotage and passive aggressively hurt those people maybe for years afterward. If you remove humans from having direct access to these situations, a lot of people are way more happy about it.
But again, not all implementations are equal. There are a lot of AI implementations that will basically outperform nearly any human court agent in customer satisfaction ratings. There are others that are so dumb that they just decrease satisfaction in the company because they are extremely not helpful and are obvious attempts to Stonewall. So to some degree you have to take into account what the objective is and the kind of implementation.
I would posit that we just need to ensure there are some general standards and society should reward those who comply and not those who donât. The incentives will decide how things go And what is more valuable and important.
We dumb down people over generations and then complain that we have no choice weâd like to use humans but theyâre too dumb ai is our only hope and people can be funded to fade away peacefully staring at their linty navels and autism can conquer the universe!
Wheeeee!
Veer is my trigger word.
More donât doozes:
I donât have the chops to define these off the top of my head. If only there were others who do who could contribute.
I know. 
@dave27âs busy. @t.currieâs peeved. and @bryonhu is too 
to stoop to peep projectz.
Alas. 
p2788deal is game, though. Maybe him?
And @K6CCC , such a shame heâs ignored me.
And @wildbill @ssummerlin @antonius @habib (donât get pouty I canât name you all. POST! 
)
@AnnWithAPlan @towelkingdom ?? 
Ten mentions to a post max what can you do???
This may be like when I invited ten people to my eighth birthday party and nobody showed. No worries, it made me stronger! 
Donât know about those things.
Been busy getting ready to go back to calif for my motherâs funeral.
No worries, Ann, take care. Thanks for stopping in.
Should say âunless you really know what youâre doingâ. Of course if you do, then you realize that QOS when you only control one end of the connection is mostly useless and often does more harm than good.
When I ran servers out of my house, it was handy for throttling individual downloads (which from my perspective was upload). That you can control. So it can be useful if you have slow upload speed like many cable providers, but again you have to know the right way to do it.
My contribution to âdonât doozesâ is mess with any setting that you donât understand or havenât researched. Applies to wyze but especially when people get into their router admin interface and start changing stuff around, then complain stuff doesnât work right. I liken it to when someone turns on a light switch and nothing happens, but then doesnât turn it off. That thought process baffles me.
Actually, contribution #2 - donât enable IPv6 on a home based router unless you have a specific need for it (most donât) and know the implications. Any understanding or knowledge you may have about IPv4 (especially in the home environment) pretty much does not translate at all into the IPv6 world.
= 
Succinct.
Is it the consensus that they might be ok for one or two nearby cams but you should upgrade the router exceeding that?
Or is it pony up for a better one and save yourself the trouble: everything you currently have will perform at least as well and thereâll be room to grow as you get more devices?
Are some people still renting ISP routers?
Yes they are, thatâs why ISPs are still offering them.
My old Motorola cable modem/router kept dropping the signals after 6 years so I installed a NETGEAR CAX30 a few months ago. It was a pain in the a$$ to set up but I finally got it set, named the network and SSID (S) the same as the old one and all is well. I have Xfinity and have never used any of their provided equipment because I donât want to pay them another $14 or $15 dollars a month and I want to be in control of all the settings. 
