It's been a year since Axios first examined the data, and as of now, there are 55 companies with self-driving car testing permits in California and 54 new accidents, based on filings of incident reports in the state.

But the one constant point is that humans continue to be the cause of most accidents, Axios' Kia Kokalitcheva reports.

By the numbers: All those beige cars in the chart above indicate incidents where autonomous vehicles were not considered "at fault" — that is, people were. Even when AVs are at fault, that's most often been in cases where humans were at the wheel ("conventional mode").

The bottom line: Unless self-driving cars magically replace all conventional cars in the country overnight, robots will have to drive alongside humans. What’s more, they’ll have to drive alongside pedestrians, cyclists and other humans with whom they share the road. And at the moment, self-driving car tech doesn't seem to be advanced enough to handle all these humans.

• This is in part why some companies are deploying their first vehicles on specific routes or defined regions, where the cars will interact with a limited set of surprises. For example, Drive.ai is starting in an office park in Frisco, Texas, while Voyage got its start giving rides to residents in retirement communities, and May Mobility is developing shuttles for company and school campuses.
• Recently, artificial intelligence expert Andrew Ng also suggested (stirring up some controversy) that it’s pedestrians who need to improve their behavior and train to maneuver safely around self-driving cars.

Yes, but: It’s still hard to compare self-driving car accident rates to those of human drivers, despite the widespread hope that self-driving cars will be much safer than human-driving ones, as University of Central Florida's Peter Hancock points out in a blog post.

One new thing: In three incidents, humans intentionally attacked a self-driving car, such as by hitting it, or climbing on top of it.

Flashback: Axios’ first look at the data in August 2017.

The tech industry was the latest target of a tweetstorm by President Trump yesterday, when he railed on Google for bias against conservatives and then followed up with comments from the White House that expanded the ire to include Facebook and Twitter.

"I think that Google and Twitter and Facebook, they are really treading on very, very troubled territory and they have to be careful. It is not fair to large portions of the population," the president said.

Economic adviser Larry Kudlow later said the White House was looking into the issue, but declined to offer more details.

History lesson: As Axios' Dan Primack points out, it was just a month ago that Trump was praising Google as a "great company" being wronged by the European Union.

• Some trace the source of Trump's outburst to a segment Lou Dobbs did Monday night on Fox Business. Others just joked that Trump googled himself and didn't like what he found.

On the other side: Whatever the origin of Trump's beef, Google rejected the notion that politics influence its search results.

"Search is not used to set a political agenda and we don't bias our results toward any political ideology,” the company said in a statement that followed Trump's tweets.

Be smart: Axios' David McCabe has a look at what Trump could do if he really does want to crack down on Google, though most actions would require an act from Congress.

Several outlets published explainers for the president on how Google actually works, including Quartz and the Washington Post.

What they're saying:

• Recode's Kara Swisher: "Thanks to Trump for yet another dopey sputter -- this time on Google -- to make the exact point of my next column for @nytopinion (plot spoiler: it's hard to make tech giants sympathetic, but you manage to pull it off with cloddish aplomb)"
• The Koch brothers-backed Freedom Partners: "Regulating the results of Google or other tech companies is a reckless idea that would undermine essential elements of free speech. Allowing the government to mandate the content that search engine providers display would set a dangerous precedent."
• BuzzFeed's Charlie Warzel: "most frustrating about Trump’s Google tweet is that algorithms are indeed biased. And they are indeed opaque. But not necessarily in the stupid, simplistic and politicized way our uninformed 5AM tweeting president suggests."

Meanwhile, the New York Times reports that more than 100 Facebook employees have joined a Facebook group called FBers for Political Diversity, after one employee's post criticized the company for having "a political monoculture" that stifles conservative dissent.

A number of Facebook users were surprised Tuesday when some of their old posts disappeared — specifically, all the updates that had been cross-posted from Twitter. It turns out Twitter was surprised, too.

The bottom line: Facebook had announced in January that it was removing a feature that allowed people to cross-post updates from Twitter. As a result, Twitter deleted its Facebook platform app, which had been rendered useless when the changes went into effect earlier this month. But it had no idea old posts would go away once the app was removed.

Here's what happened, according to a source close to Twitter: The social media platform had initially asked Facebook for more time to see if there was a way for users to continue joint posting to both social networks, but Facebook said no.

• As a result, the Twitter app for the Facebook platform was essentially made useless earlier this month once Facebook officially removed the ability to cross-post.
• With the app's sole function eliminated, Twitter decided to delete it from the Facebook platform, having no reason to think that doing so would remove old tweets that were cross-posted. It's not clear whether Facebook knew this would happen, either.
• That said, the content has apparently now been restored.

“A Twitter admin requested their app be deleted, which resulted in content that people had cross-posted from Twitter to Facebook also being temporarily removed from people’s profiles," Facebook told Axios late Tuesday night. "However, we have since restored the past content and it’s now live on people’s profiles.”

Why it matters: The deletions were brief, but the snafu serves as a reminder that it’s not always clear who has control over user data on giant social platforms — and it’s often not the user.

Three reports out Tuesday suggest that Amazon is pushing harder to corner the entertainment sector, with plans for a free TV streaming app, a live television recorder and a new slate of content deals with two of Hollywood's biggest movie studios.

Why it matters: One of Amazon's fastest-growing business units is advertising, which will continue to explode with investments in content and TV technology. Content investments will also help the company take on Netflix in a heated battle for digital streaming subscribers, reports Axios' Sara Fischer.

Three new investments:

1. Amazon is planning an ad-supported Fire TV streaming app, nicknamed "Free Dive," that would rival that of Roku, The Information reports. Roku currently offers a free, ad-supported app for Roku devices called "Roku Channel." The Information reports that Amazon, like Roku, is in talks with major studios to license older TV shows to air on the app.
2. It's also talking to Sony and Paramount for streaming rights to add more programming to its Amazon Prime video service, which is different from the planned "Free Dive" app, Bloomberg reports. (The app is for only for owners of Amazon's connected device, Fire TV.) Bloomberg also says Amazon is talking to Sony's studio division and Viacom's studio Paramount Pictures.
3. Amazon is also planning its own DVR service that would rival that of TiVo, Bloomberg reports. The device, nicknamed "Frank" within Amazon, is a new type of digital video recorder for the streaming era. The device would include physical storage and would connect to Amazon’s existing Fire TV boxes.

Read more of Sara's full story here.

Go deeper: How Amazon is eating the media business.

On Tap

• Salesforce reports earnings after the markets close.
• VMworld continues in Las Vegas.

Trading Places

• Marketo hired former SAP and Microsoft executive Mika Yamamoto to be its global president.
• JP Morgan nabbed Google's Apoorv Saxena to run its AI efforts. Saxena had been head of product management for cloud-based AI. (CNBC)

ICYMI

• TV providers Altice and Dish Network want the U.S. government to block T-Mobile's proposed purchase of Sprint. (Reuters)
• Chinese EV maker NIO is seeking to raise up to $1.3 billion from a U.S. IPO that would give it a valuation as high as $8.5 billion. (Axios)
• California legislators are considering a first-of-its-kind bill that would require publicly traded companies based in the state to have at least one woman on their board. Some European countries, including Germany, have such gender-based mandates, but opponents say that the government has no authority to do so here. (Axios)
• Box beat quarterly earnings estimates, sending shares slightly higher in after-hours trading. (Reuters)
• An apparent Iranian influence operation targeting internet users worldwide may be significantly bigger than previously identified. (Reuters)
• The recent state of tumult for CEO Elon Musk and Tesla is spurring fresh chatter about a question that has been rattling around for a while — should Apple buy Tesla, or at least acquire a major stake? (Axios)

The Lego minifigure just turned 40. Happy birthday!