Telegram user here. I'm not anti-bot but puh-lease, spend this money on audited open-source crypto and secret-by-default chats rather than gimmicks like 'stickers' (yuk) and rewards for promising bots.
I want a messaging app that's simple, secure, open and cross-device. Not one that's splurging on daft bloatware features while the key issues go unaddressed.
[edit] to clarify the term 'cross-device' - I mean I want to see my conversations on my phone, tablet and laptop and easily swap from one device to another. Platform support is a separate, also-important issue that Telegram already has nailed.
I figured someone would say that. I want actual cross-device, not Signal's weird "link your phone up to your MacBook" solution. So not Signal as it presently stands.
Also (last time I checked; maybe it has improved) Signal's support for inline media within conversations was light years behind Telegram.
Don't get me wrong, I applaud Signal's focus on security, but its usability simply isn't there yet.
I find signal really very usable. Huge doubts over telegram's security - it isn't there yet and doesn't sound like they even want it to be. Telegram seems like noise if security is interesting to you.
I can't reply to the post underneath yours but I very much agree with you and for me the reason I don't want the "link your phone to your macbook" solution is because my phone and it's number are very much not going to be permanent.
I do however have an email address I intend to use for a very long time.
I was actually looking for a chat solution earlier today and neither telegram nor signal fit the bill for these very reasons.
I want telegrams nice-looking apps, with signals security. Could I use WhatsApp? Maybe, but Facebook owning it makes me uncomfortable.
Why aren't you using Wire (https://www.wire.com)? It is cross platform, end to end secure w/ open source audited crypto, focuses on app usability, and doesn't need a phone number to sign up.
Trust issues with the founders (Skype hardly had a great privacy record); inability to build own APK (because only the crypto is open source) means you can't be sure the code on GitHub is what you're really running.
I had honestly never heard of it before. It looks good, but the website is completely lacking any information about the company which again makes me feel a little bit uneasy. I might try it, though.
> I want telegrams nice-looking apps, with signals security. Could I use WhatsApp? Maybe, but Facebook owning it makes me uncomfortable.
Exactly, bots are an interesting and cool idea, and they're obviously putting a lot of effort behind it. But it just pales in comparison to the huge leap WhatsApp took, by implementing Signal's encryption.
TBH, I hadn't expected that move from WhatsApp at all--and I'm still a tiny bit skeptical, because it almost sounds a bit too good to be true :)
I have been using Signal for a bit, too, but hadn't gotten around to read up on how Signal's encryption protocol works, but I intend to. This too seems almost too easy, to get zero configuration and hard security. Surely there must be some pitfalls or trade-offs I'm paying for that? Ok there's the verification codes, and I can see if the protocol properly tight and secure, that might get you there. But I expect that to still leave some minor pitfall(s), and I just kind of want to know what exactly those are.
On the other hand, Telegram's Linux desktop client is just glorious. Sync is perfect and instant. The whole GUI runs super-snappy even on my 4-year old low-powered netbook. Which is something I can't say for WhatsWeb. Because it's so smooth and instant-sync I use as one of my primary means of sending text or data between my netbook and my phone.
The other reason is that, my data is being slurped left and right by big foreign corporations operating under very nosy governments. Currently that is mainly the US government, given the software platforms I use. Even if I can't help those powers tapping my (meta)data, it still makes a lot of sense to divide up my data-trail between powers that are very unlikely to share this data with eachother.
And then, about half of my friends are on Telegram. Many of them for the two reasons above, others because the rest is there :-P I personally don't mind using more than one IM-app/network either. I have all the "messaging" stuff (including GMail, btw) grouped in one of those "folder" like icons on my homescreen (Android/Cyanogenmod), giving very quick access to any I need.
So yes, Telegram. Make with the hard encryption. I personally prefer it over bot-functionality, even though I think it is very shiny. But what I would really hate to see is an IM network with mediocre encryption "win" (popularity) over a network with much better encryption, just because it has a shiny bot platform.
So use chromium? Or have your messages stored by a third party in clear text - which seems like a far worse option than having a browser you don't otherwise have to use at all, installed.
No, you missed the point. They don't want Chrome, and if their aversion could be solved with small tweaks to the branding they'd have probably just gotten over it.
If you don't want to use chrome or chromium to run one single app to have proper encryption on your desktop in addition to your mobile device - which doesn't need a browser at all, and as a result of that "aversion" would prefer your messages are kept in clear text on a server your aversion is not worth considering further.
Or I'm wrong, what's the aversion that is more important than having your messages stored in clear text by a third party?
If Signal was available on F-droid and didn't depend on Play Services and Chrome (for Desktop) I'd be happier.
In my opinion Open Whisper Systems sacrificed the power users to be a safe messenger for the masses when they could've pleased the power users, who are key to driving mass adoption, too.
You make it sound as if people who do not want to use Play Services are the only power users. That is simply false, many power users don't care about using Play Services.
I think you're missing the point. Bots now use ML and other forms of weak A.I. Facebook is building something closer to artificial general intelligence with Facebook M. The early forms of bots we'll see will do basic tasks but in the future they'll be as smart or smarter than us. This is a great long term play Durov took by incentivizing bot developers. Besides, Telegram doesn't have the same scale as Facebook (FB Messenger) and Microsoft (Skype) so they sorta need to do this to remain competitive in their hyper competitive market.
If I had any gripes about Telegram it's their lack of audio and video calling but I'm not going to complain about an excellent free product.
Don't discount stickers. Telegram's sticker support (and its ease of creating your own sticker sets) is one of the reasons I and a lot of my friends have moved to it for most of our messaging lately. I personally went crazy and did a set of like 50 stickers with one of my own characters, which can add a ton of whimsey to my casual conversations.
Telegram really doesn't seem like they're trying to be a secure chat client at all at this point. It seems like the focus is more on making a follow-up runet social network now that Durov no longer has control over VK.
It echoes VK in a lot of ways (pirated media abounds!), but channels are some weird cross between Twitter and 4chan as opposed to being a blatant Facebook clone. Direct messaging is still there but it doesn't seem like the network is focused on that at all anymore.
Underneath the original "wow factor", I think bots have a glaring UX problem. Developers just haven't figured out the real use-case for bots. In some ways, I think this current bot-craze is very similar to what happened with smartwatches 2 years ago. Simply cramming a phone app into a watch doesn't make it magically awesome or useful.
I think a lot of developers seem to be taking the same approach with bots. They are trying to create alternative UIs for their already existing mobile apps through bots. Mobile / web apps do a lot of the same stuff as bots in a much more user-friendly way. Take a look at https://storebot.me/ and tell me if there's one bot that can do something better than a mobile app.
I'm not anti-bot but I just think the bot-frenzy is premature,
I think that this is a better platform than mobile as it supports multiple devices, user accounts are already setup, it's easy to test and it has built in notifications.
One thing I did have an issue with, is showcasing what it's features were. One option is to create a lot of videos and screenshots which would take a lot of time. As further improvements were made, I would have to reshoot the videos.
Nice! Yeah I think it has potential for specific kinds of applications. Text-based games are a definite yes. I couldn't agree more with your point about user education. I know a lot of friends who don't even know all the amazing things Google Now can do just because there is not interface to explain all the features. Siri has a "what can you do?" feature but even then, it's very limited. It will be interesting to see what developers and UX professionals do to improve the state of bots.
What money? There's no platform for payments or advertising on Telegram. Advertising doesn't seem very practical with bots. Telegram has given no details or examples of what qualifies for the $25k rewards. It's not clear what level of production they're looking for... a bot that tells you the weather or a virtual girlfiend with full-blown NLP? Innovation prizes are cool but you can't play if you don't remotely know what qualifies as winning.
What about API library developers [1]? We haven't received anything yet but thanks to us others can build bots. I invented the Telegram bots back in 2014 [2] (there wasn't API) and haven't received any thanks from Telegram.
Would Telegram let you develop a "bot" that supports connections to other messaging systems? Or are they into the walled garden with spikes on top of the wall thing?
If they have $1M to invest, why don't they focus on securing their protocol (i.e. switching to authenticated encryption) and turning encryption on by default?
I'm not up to date, but didn't Telegram had a bounty for someone who could read a conversation even by knowing the phone numbers involved? If so, why the writer here didn't got the bounty?
So telegram is a non-profit. And they splurge 1M USD? What am I missing?
Edit: to clarify, what I meant is that I expected non-profits to make good use of their money and not give away large sums of money in prize competitions. Certainly, no NGO or non-profit I know of does this. Raises a red flag for me.
non-profit does not necessarily mean the company does not make profit. AFAIK it just requires that the profit made be spent on the company and/or cause rather than used as dividends or given to what would normally have been shareholders.
Are you referring to where I wrote "profit made be spent on the company"?
If so, I'd like to point out that expenses are not the only way you can spend on a company. Someone may also invest profit that they have made into expansion, for example.
"profit" can never be consumed by expenses, by definition (as you have pointed out). Profit can, however, be re-invested in a company, as opposed to being taken by the owner or shareholders, or whatever.
How I'm understanding is that these things happen at different times.
1. Business operates for X days (1 quarter, 1 year, whatever)
2. Business calculates it's profit (revenue - expenses)
3. Business now has a choice on what to do with it's profit.
At t = 3, the profit is not an expense, though it will ultimately be used to benefit the company/cause (if it's a nonprofit).
nonprofit organizations must decide what to do with profit, if they have generated more revenue than they have expenses. The difference between a nonprofit organization (from another org) is that the profit must go either back into the company, or into the cause. Arguably, the money that goes back into the company goes into the cause anyway, as it is in service of the cause.
Maybe it's just a matter of semantics, I could see it being incorrect to call something "profit" that is never allowed to be spent on anything other than the company/cause. However, the strict definition of profit is pretty bare-bones, simply revenue - expenses, so I think it qualifies.
A nonprofit with profit is a contradiction. I get that some may perform calculations with labels called profit, but that contradicts the underlying thesis of the organization (as you've admitted).
No, that doesn't explain anything. Durov has a lot of money to throw around. Okay. Why did he throw it at something silly when Telegram has real problems? NCC Crypto Services would do a lot of good work for a million dollars.
The original comment asked why a non-profit would splurge $1M on this. It's not the non-profit's money though, it's Durov's own, so his net worth is totally relevant.
By now I haven't heard anything of how they make or will make money. It's like nobody addresses that. I know a messaging app is not exceptionally expensive to run but still. Anybody has any hints?
There are other nonprofits that do similar. The XPrize, for example, has similar contests for space travel. Contests like this encourage groups that may be happily in the shadows to come forward and be bold, and show off what they've been working on. The money definitely will help, but creating names can be more worthwhile.
@watchbot is a useful bot for monitoring uptime, but I am afraid that it also probably gives FSB my uptime information, and I do not have control over that.
So many negative comments, would like to see what happens if exact same title would be with Facebook, considering their scale lets say 'Facebook: $10Mln to Bot Developers. For free.'.
Do not forget, they may have some issues with security or user growth, but founder generally tries to promote his own product, everyone will behave like he does (promote in product in any legal way), when they have enough resources.
For academics, when you get a fellowship or grant that must be recorded on the 1099-misc in the united states. And you pay full taxes as an independent contractor.
I am not a lawyer nor am I an accountant. This is not legal or tax advice :)
Good news, everyone! This year, Pavel Durov will be giving away one million dollars to bot developers in the form of grants starting at $25,000 each.
Note that 'grants' is bolded in the article. A grant is not a prize, and a prize it not a grant. So it's one or the other, and since what they are doing would normally be called a prize (or reward), the fact that they specifically mention grants is likely because it has special significance.
It's like if I say to my uncle, "I'm giving you a gift of $10,000 in the form of a loan. Enjoy your $10,000". A gift can't be a loan, and a loan can't be a gift (for tax purposes). But I might say that, because the tax treatment for loans and gifts are different.
So that goes to my question of whether or not calling it a grant has tax significance.
Why not allow charging for bots and stickers? It stifles bot innovation as there are many great bots that have real costs that must be sustained somehow. Same with stickers. Great artists have zero incentive to make sticker packs since they can't charge for them and anyone can give them away to other users.
Developers and artists should be treated like stakeholders rather than cannon-fodder and free labor.
Can someone explain to me what bots are for. It looks like using something similar to a REST API over Telegram. Whats the point / use / advantage of that?
We at Microsoft think about bots as somewhat "intelligent", as in - they understand your intent. There's a little example chat with a pizza bot: It's a subtle difference, but being able to have a conversation vs navigating a flow chart is pretty powerful for a bunch of applications.
I wonder if you had a look at what I consider your largest competitor, Facebook Messenger. What are the pros and cons of wit.ai compared to LUIS, in terms of quality, performance, ease of setup?
Bots offer a gateway between human language and API actions. The future (which VC's are very excited about) is one where the human language commands evolve into something non-syntax specific.
- loose syntax, as in human language, which we can parse although it is loose (e.g. I went to the shop, I went to shop, I went to de shop - all are parseable by humans to get the same meaning [1])
OR
- something else?
[1] except that "I went to shop" could also mean "I went shopping" - but ignore that for now, it's just not a good an example.
Anything the bot-builder wants really. I think the aim of this grant is to explore the limits of what bots could be used for.
Bots don't have to use AI or ML to be useful: right now I use the @gif bot to search for and reply with reaction gifs inline (without having to launch browser->search->download->attach into chat). You can create a Jenkins/CI bot that chimes in when a build fails, or create a weather bot that tells you it's going to rain
I want a messaging app that's simple, secure, open and cross-device. Not one that's splurging on daft bloatware features while the key issues go unaddressed.
[edit] to clarify the term 'cross-device' - I mean I want to see my conversations on my phone, tablet and laptop and easily swap from one device to another. Platform support is a separate, also-important issue that Telegram already has nailed.