It always puzzpes me that people with zero idea how dirt cheap shipping is per unit believe drones, or moving production closer to consumers, woupd ever work from a price/cost point of view... And since the last mile is the most difficult one, supermarkets are pretty good in outspurcing that to customers.
For now. Who is going to be the first company to pull it off though? I'd definitely subscribe to that - would love fresh produce parcels dropped off on my front steps or backyard by drone delivery. This has already been done with medical supplies before which is super cool.
I really don't think it's feasible without some kind of AGI (or human) on the controls. There are so many environmental, technical, societal and regulatory issues to tackle. They make a lot of noise (although some are much quieter now), people could very easily jam their communication signals (or just shoot at them), birds, hardware failures, ... I worked on last-mile delivery project at a drone startup, we could do cool demos that impressed shareholders, but I could never see that working in the wild, people would get hurt.
Also, once you solve the last-mile problem, a new one arises: the last-foot problem...
Make it available as an open source platform and I’d want one.
We need to know exactly what software is running on it and what the hardware is, though. The user must have complete control over the hardware, software, and connectivity at all times.
It will all be alright. This is just an understandable and expected recalibration for the global economy after a global pandemic, global supply chain crunch, and the Ukraine-Russia war.
The United States and Europe are fundamentally rock solid and poised for considerable growth and innovation for a long time yet, along with their allies and friendly rivals.
I tend to agree. Fundamentally, we're pro-market, liberal democracies. Compare that to Russia and China, for example. Those two are heavily controlled, corrupted plutocracies or even dictatorships. For all the issues America and Europe have (and we do have issues), I think the pillars of our economies and our societies are much better.
This seems like a past performance as a measure of future returns error to me.
China has had a lot of success with free market and exports while having substantial central control and without being a liberal democracy. It's not really clear how systems will fail or succeed as technological balance rebalances or unbalances power between the estates.
I think long term leaders will continue to be parasitic burdens upon their societies but its quite possible that technology will help alleviate their typical consequences on productivity.
For example, virtually every corporation is at its heart a failure in democratic control and a centrally controlled institution with either a long term dictator or oligarchy. They had substantially more problems competing in earlier times when paperwork overhead was literal paper.
For me it was reddit's api. That being said, they can't be expected to provide free or low-cost apis across the board forever...otherwise everyone's gonna just constantly use them to the detriment of the entire platform! There has to be a cost so that an equilibrium between demand and availability is attained.
Tbh, $100 a month for a hobby project or prototype is not the end of the world. Maybe they could add special student pricing for student hobbyists. Aren't they also still doing a free option that's reasonable for testing out the api?
Yeah, but if the bulk of the world's population is slamming an api that's free...there's no more api or platform at that point, since it's not going to generate enough revenue anymore. Could take the entire platform with it if all of the available funds are going towards serving free api traffic.
I agree with what you're saying at a high level - there's no such thing as free lunch forever, but I disagree sharply in terms of cost/scale necessary for profitability.
$100/mo gets a hobbyist:
- Low-rate limit access to suite of v2 endpoints
- 3,000 Tweets per month - posting limit at the user level
- 50,000 Tweets per month - posting limit at the app level
There's no world where that costs $100, $10, or even $1. My GoogleFi plan costs me $10/1GB.
It's wildly apparent that Twitter is overstating the value of the service they provide to the detriment to their users. They aren't trying to offer break-even prices for hobbyists.
There's more to posting/saving a tweet than just the data sent over the wire. There's many indexes, notifications and other analysis that happens. The free tiers of the API usage are what was bringing Twitter to it's knees in terms of overhead, not the actual users.
And bots using a browser with actual interaction patterns that at least resemble a person (not pulling through hundreds of messages in under a half a second or posting in 10ms after the form loads) is much more of a throttle, especially with per-ip rate limiting in place than the raw API access.
> Could take the entire platform with it if all of the available funds are going towards serving free api traffic.
You seem to have a really skewed sense of how Twitter's API actually works, and you also seem to be under the incredibly incorrect impression that the API has no rate limiting.
Like most services, Twitter actually saved money by offering a free API with OAuth, because the alternative is for people to use web scrapers and direct access by password, which is orders of magnitude more expensive both due to direct network traffic costs and due to the security costs of people/apps doing insecure things to get around the lack of an API.
> Yeah, but if the bulk of the world's population is slamming an api that's free...there's no more api or platform at that point, since it's not going to generate enough revenue anymore.
This is literally an argument for requiring all Twitter users to pay to use the service, since the official apps all use the same API.
From a systems engineering point of view you want people hitting an API for programmatic requests. The alternative isn’t “The bots magically go away”, they just get more sophisticated, mix in with and impersonate user traffic, and become impossible to properly rate limit without collateral damage. APIs are for the benefit of the service provider as much as they are for developers.
This misses the point that the op is making: $100/minus cost prohibitive for anybody looking to use the API casually or explore deeper integrations for it.
You might be able to get a solo developer to pay $5 - $20 USD per month to trial something, but few will be willing to drop $1.2k/year for marginal gains.
I honestly think they don't care if they turn off developers. The core of their platform is advertising, and API users don't see the ads. I'm also guessing that API usage dramatically outweighs actual users in the past. And while I'd rather see an ad free paid tier, I get it... they were dramatically over-valued and the expenses needed to be reigned in from where they were. It was unsustainable.
I'm not sure that I understand the downvotes... Unless the suggestion is the twitter management absolutely cares about small/indy developers or that they didn't need to cut expenses.
Because its a decision made by someone who clearly has no idea how modern web traffic works. That pricing model looks fine in a world where scrapers and proxies don’t exist. But in our world, they do exist, and small time devs will happily pay $5 for a handful of proxies and scrape the data they want from Twitter directly. You can’t really block them either, especially if they’re indeed small time devs who are just scraping casually.
These devs using scraping tools are loading up entire profiles and tweets, which consumes far more resources than a simple API call that gives them the precise information they want.
Cheap APIs discourage scraping and are most cost effective if you work out the tiers and rate limits.
And my point is that they don't care about the $5/month devs. And the scrapers or intermediaries will just expend their own resources. With less overhead on Twitter's backend.
It's pretty easy to throttle browser requests without anyone noticing and blocking excess requests from a single IP block.
That you think they should care, doesn't mean they actually do.
> That being said, they can't be expected to provide free or low-cost apis across the board forever.
They did, for most of Twitter's history
> ...otherwise everyone's gonna just constantly use them to the detriment of the entire platform!
On the contrary, the free availability of the Twitter API is inarguably what drove the growth of Twitter as a platform. Twitter benefited far more from it than the low marginal costs of operation. It's not even close.
I pay $200 a year ($17 a month) for access to a large, productive portion of VMWare's datacenter stack through VMUG. 6x that for access to a social media API is a terrible deal. At $100/mo, Twitter is basically saying "We don't want your business".
That is more likely the case here. They don't see profit from their side to make lower level access between the free and $100/mo tier as worth it... and would rather see corporate backed developers who are likely to have 5+ figure annual contracts than deal with tens of thousands of $10-20/mo subscribers disproportionally using the system.
Those $100 would have been barely enough for testing some things. That's more than my spending limit has been as 13yo.
Sure I get what you mean, and I couldn't care less about today's twitter.
But it still makes me sad to know that in today's work. 13yo me without PayPal or credit card wouldn't have any of that fun. And I imagine millions of people can't either.
The $100/mo API has extremely shit limits on top of being prohibitively expensive for prototyping. This will kill the mere idea of building anything on Twitter's API. Low friction and upfront costs are important for hobby projects and experimental prototypes.
In this day and age it's expected that companies have generous free tiers for hobby/prototype usage, and it makes sense because it usually costs them pennies to provide it.
Serving API requests is extremely cheap and the volume is almost always going to be so insignificant that they wouldn't even be able to tell if developers are using the API if not for API keys.
> they can't be expected to provide free or low-cost apis across the board forever
They _could_ if it supported people building stuff that increased engagement so they could sell more ads.
I think it's mistaken to think of free-to-access APIs as something the company 'gives' to users at some cost to itself. Done well, it facilitates making the platform a richer place, where people spend more time and attention, and which is therefore more valuable. It's hard to do the attribution to definitely say k% of timeline views (and thus ad impressions) wouldn't have happened without API-dependent stuff, but that doesn't mean it's 0%.
Some categories of examples of stuff that I think formerly contributed to engagement but which would just not get built today:
- write only twitter-bots which give information on e.g. earthquakes, public transit delays, in a way which is genuinely informative, and does not enrich the author(s).
- interactive twitter bots which made twitter itself better to use. The most important in this category may have been Threader, which was ultimately acquired by twitter. But it would have been useless / never written under the new rules. Can you imagine trying to call it, and receiving no reply b/c it had exceeded its limit for the day?
- interactive twitter bots which made twitter a platform from which to do other stuff. E.g. your.flowingdata.com was a self-tracking project where you recorded information by tweeting. Treating twitter as a platform, and orienting itself around tracking routine stuff meant that using this project _required_ you to frequently engage with twitter. This was also a free offering, which wouldn't have existed under the current limits.
An ecosystem in which there are high costs to building means less stuff will be built, and the platform overall is less interesting, less compelling, less worth scrolling through an ad to see. Thinking of these APIs as just a cost center is misguided.
Then they could provide it for a year, or six months, or one month.
twitter currently makes it's money from advertising served on content users generate, creating barriers around programmatically serving that content reduces the value of the platform for advertisers.
twitter is, I guess, looking to switch to a subscription based model, but paid subscription models make the most sense when the content is exclusive in some sense, and create barriers for entry, especially for folks from outside the wealthier nations.
Maybe we just wrap planet Earth in more tech so that it becomes a ship and we can steer the entire planet, maintaining a climate with tech. Risky, but once we have more planets settled we will be able to experiment with this.
...Is a book I have not read, but the author is also the author of my favorite non-parody Scifi series (Remebrance of the Earths Past / Three Body Problem).
