Google engineers quit after the company fired its star AI ethics researcher

Two Google engineers have quit the company over the firing of AI researcher Timnit Gebru, whose controversial termination in December sparked an internal backlash about ethics and diversity.

David Baker, who led the engineering team for Google Trust and Safety, said he was resigning after more than 16 years because Gebru’s departure had “extinguished [his] desire to continue as a Googler.”

Software engineer Vinesh Kannan has also left the company. He tweeted that Google’s treatment of Gebru and diversity recruiter April Christina Curley “had crossed a personal red line.”

Gebru is a widely-respected AI ethicist, who’s best known for a landmark study on race and gender biases in facial recognition systems.

She lost her job at Google after the company asked her to retract a research paper about the risks of large-scale language models — which are used in many of the firm’s products.

Gebru then sent an email to colleagues expressing her frustration with the response and the company’s wider issues around diversity.

In a letter dated January 5 that was posted to LinkedIn, Baker said her exit had undermined his efforts to build trustworthy products and support diversity.

Curley has been a prominent public defender of Gebru and critic of Google’s diversity efforts.

In December, she said she was fired because the company was “tired of hearing me call them out on their racist bullshit.”

Baker cited the company’s lack of diversity in his resignation letter:

The resignations of Baker and Kannan come amid mounting internal dissent about Google’s AI work.

Thousands of employees and external supporters have signed an open letter expressing solidarity with Gebru and calling for Google to honor the commitments made in its AI Principles .

The new Alphabet Workers Union also cited Gebru’s exit as one of the reasons for its formation.

Google CEO Sundar Pichai has attempted to subdue their anger by pledging to investigate her departure and apologizing for it causing staff  “to question their place at Google.” But the tensions continue to escalate.

Theory: We’re running out of time if we ever want to meet aliens

It seems like just yesterday the dinosaurs were frolicking about the canyons of Pangaea, living their last few oblivious moments before an asteroid changed everything. You don’t really notice it while it’s happening, but 65 million years goes by pretty fast.

At least it does in the grand scheme of things. The universe has been around some 14 billion-or-so years according to the Big Bang Theory. In that context, 65 million years is just a few flakes of sand in a giant hourglass.

And that makes it all the more scary to read a pre-print research paper suggesting that our universe may only have a measly 65 million years of expansion left.

Once that expansion ends: it’s lights out. The universe will transition to a (hopefully) slow contraction until it becomes a single infinitesimal point made up of all the matter that’s ever been and ever will be.

At this point, according to scientists, the Big Bang will happen. Again.

A broken clock

The truth of the matter is that we have no way of knowing if and when the universe is going to end. It’s not that we don’t have the technology. We don’t have the perspective.

When we look up at the stars we’re gazing at light that’s traveled for millions or even billions of years. Some of those stars don’t even exist anymore.

Yet, by observing these distant lights we’re able to determine that the universe is expanding. And, in a series of simulations, scientists have used that data to try and figure out whether that expansion is infinite or if there’s some sort of universal reckoning coming.

To that end, a team of scientists from Princeton and NYU recently published a pre-print research paper describing the problem. According to them, if dark energy is responsible for the universe’s expansion, it’s plausible the energy could wind down and lead to a contraction.

Per their paper:

Basically: the universe gets big, then it gets small.

However, as mentioned above, we’re in no position to know when this will happen. All we can do is guess.

According to the scientists:

And that means, for all we can tell, the universe could continue expanding for billions of years. Or, perhaps, the time is already nigh.

The Physics arXiv Blog describes the researchers’ time-frame as such:

The end of days

Right now, that 65 million years time-frame is just a guess. But if we could definitively determine that our universe has such a short amount of time before it starts contracting, the ramifications of this knowledge could have immediate, far-reaching impacts on humanity.

It would essentially mean we’re unlikely to ever find alien life. This simply boils down to the odds: if we look at 65 million years as the final few flakes in our universe’s hourglass, we have to concede that time’s almost up for all living beings.

If we haven’t found each other by now, the odds are mathematically against it ever happening.

Between expansion spreading us apart and contraction ultimately resulting in everything in the universe being compressed to a single point, it may even be pragmatic to give up the search all together and focus on something else. We could use our space resources to advance the goal of spreading Earth’s life throughout the galaxy, for example.

The universe may only have a little time left, but it’s worth making sure we’re all there to watch the grand finale. That is, of course, unless the world really did end in 2012 .

The scientists also postulated that the universe’s expansion and contraction were cyclical, meaning another Big Bang is likely imminent.

On the bright side, this would mean there’s almost certainly going to be life again once we’re gone. But, on the not-so-bright-side, it also makes it less likely that life is prolific.

With such a tiny window for life to miraculously appear during each expansion and contraction cycle, planets like Earth might be once-in-a-Big-Bang occurrences.

Nope, ‘facial recognition’ didn’t spot Antifa members at US riots

After right-wing rioters stormed the US Capitol building on Wednesday, Republican Congressman Matt Gaetz pushed a conspiracy theory that members of Antifa masquerading as Trump supporters were among the mob.

Gaetz, an ardent Trump ally, based his claims on a Washington Times report containing “compelling evidence” from a facial recognition firm.

The story features a cast of shady characters: a controversial Congressman who’s been accused of glorifying violence , a pack of far-right protestors; a publication with a history of shoddy reporting , and a mysterious facial recognition company.

But the company in question — XRVision — told TNW that the claims are nonsense. In a statement, XRVision said it hadn’t identified anyone from Antifa — although it had detected two members of neo-Nazi organizations and another individual with a history of promoting QAnon:

XRVision said its attorney had instructed the Washington Times right-wing outlet to retract its claims and publish an apology. The conservative outlet has now issued a correction and apology for the error.

XRVision has also provided sample images of the individuals its technology purportedly detected. But the company’s investigation also requires further scrutiny.

OneZero’s Dave Gershgorn notes that XRVision hasn’t published research publicly, provided online information about its technology or clients, or submitted its algorithms for testing by the National Institute of Standards and Technology.

L ike the Antifa conspiracy theory, the words of Congressman Gaetz, and the apparently unsubstantiated Washington Times report, the facial recognition firm’s claims shouldn’t be taken at face value.

Leave A Comment