Group items tagged

Ambiguities are often reconciled with other observed data, such as that we know by other means that the planet is too small to cause the scale of distortion seen.

“The two previous theories of degeneracy deal with cases where the background star appears to pass close to the foreground star or the foreground planet. The AI algorithm showed us hundreds of examples from not only these two cases, but also situations where the star doesn’t pass close to either the star or planet and cannot be explained by either previous theory,” said Zhang in a Berkeley news release.

But without the systematic and confident calculations of the AI, it’s likely the simplified, less correct theory would have persisted for many more years.

As a result — and after some convincing, since a grad student questioning established doctrine is tolerated but perhaps not encouraged — they ended up proposing a new, “unified” theory of how degeneracy in these observations can be explained, of which the two known theories were simply the most common cases.

“People were seeing these microlensing events, which actually were exhibiting this new degeneracy but just didn’t realize it. It was really just the machine learning looking at thousands of events where it became impossible to miss,” said Scott Gaudi

But Zhang seemed convinced that the AI had clocked something that human observers had systematically overlooked.

Just as people learned to trust calculators and later computers, we are learning to trust some AI models to output an interesting truth clear of preconceptions and assumptions — that is, if we haven’t just coded our own preconceptions and assumptions into them.

The simple truth is that building something valuable — whether that’s knowledge, experience or a network worth participating in — is really, really hard. But tearing it all down is piss easy.

It almost doesn’t matter if this is deliberate sabotage by Musk or the blundering stupidity of a clueless idiot.

This way of doing business rewards huge Internet platforms, since ads that are worth so little can support only companies with hundreds of millions of users.

Ad-based businesses distort our online interactions. People flock to Internet platforms because they help us connect with one another or the world’s bounty of information — a crucial, valuable function. Yet ad-based financing means that the companies have an interest in manipulating our attention on behalf of advertisers, instead of letting us connect as we wish.

Many users think their feed shows everything that their friends post. It doesn’t. Facebook runs its billion-plus users’ newsfeed by a proprietary, ever-changing algorithm that decides what we see. If Facebook didn’t have to control the feed to keep us on the site longer and to inject ads into our stream, it could instead offer us control over this algorithm.

Many nonprofits and civic groups that were initially thrilled about their success in using Facebook to reach people are now despondent as their entries are less and less likely to reach people who “liked” their posts unless they pay Facebook to help boost their updates.

What to do? It’s simple: Internet sites should allow their users to be the customers. I would, as I bet many others would, happily pay more than 20 cents per month for a Facebook or a Google that did not track me, upgraded its encryption and treated me as a customer whose preferences and privacy matter.

Many people say that no significant number of users will ever pay directly for Internet services. But that is because we are misled by the mantra that these services are free. With growing awareness of the privacy cost of ads, this may well change. Millions of people pay for Netflix despite the fact that pirated copies of many movies are available free. We eventually pay for ads, anyway, as that cost is baked into products we purchase

A seamless, secure micropayment system that spreads a few pennies at a time as we browse a social network, up to a preset monthly limit, would alter the whole landscape for the better.

we’re not starting from scratch. Micropayment systems that would allow users to spend a few cents here and there, not be so easily tracked by all the Big Brothers, and even allow personalization were developed in the early days of the Internet. Big banks and large Internet platforms didn’t show much interest in this micropayment path, which would limit their surveillance abilities. We can revive it.

If even a quarter of Facebook’s 1.5 billion users were willing to pay $1 per month in return for not being tracked or targeted based on their data, that would yield more than $4 billion per year — surely a number worth considering.

Mr. Zuckerberg has reportedly spent more than $30 million to buy the homes around his in Palo Alto, Calif., and more than $100 million for a secluded parcel of land in Hawaii. He knows privacy is worth paying for. So he should let us pay a few dollars to protect ours.

In early 2013, a few months after the mass shooting at Sandy Hook elementary school, a Yale psychologist created an experiment to test how political bias affects our reasoning skills. Dan Kahan was attempting to understand why public debates over social problems remain deadlocked, even when good scientific evidence is available. He decided to test a question about gun control.

Then Kahan ran the same test again. This time, instead of evaluating skin cream trials, participants were asked to evaluate whether a law banning citizens from carrying concealed firearms in public made crime go up or down. The result: when liberals and conservatives were confronted with a set of results that contradicted their political assumptions, the smartest people were barely more likely to arrive at the correct answer than the people with no math skills at all. Political bias had erased the advantages of stronger reasoning skills.

The reason that measurable facts were sidelined in political debates was not that people have poor reasoning skills, Kahan concluded. Presented with a conflict between holding to their beliefs or finding the correct answer to a problem, people simply went with their tribe.

It wasa reasonable strategy on the individual level – and a “disastrous” one for tackling social change, he concluded.

But the biggest distortion in the gun control debate is the dramatic empathy gap between different kinds of victims. It’s striking how puritanical the American imagination is, how narrow its range of sympathy. Mass shootings, in which the perpetrator kills complete strangers at random in a public place, prompt an outpouring of grief for the innocent lives lost. These shootings are undoubtedly horrifying, but they account for a tiny percentage of America’s overall gun deaths each year.

The roughly 60 gun suicides each day, the 19 black men and boys lost each day to homicide, do not inspire the same reaction, even though they represent the majority of gun violence victims. Yet there are meaningful measures which could save lives here – targeted inventions by frontline workers in neighborhoods where the gun homicide rate is 400 times higher than other developed countries, awareness campaigns to help gun owners in rural states learn about how to identify suicide risk and intervene with friends in trouble.

When it comes to suicide, “there is so much shame about that conversation … and where there is shame there is also denial,”

When young men of color are killed, “you have disdain and aggression,” fueled by the type of white supremacist argument which equates blackness with criminality.

Pandu Nayak, a vice president for search who leads the company’s search-quality teams, told me that snippets are designed to be helpful to the user, to surface relevant and high-caliber results. He argued that they are “usually an invitation to learn more” about a subject

“It’s a strange world where these massive companies think they’re just going to slap this generative slop at the top of search results and expect that they’re going to maintain quality of the experience,” Nicholas Diakopoulos, a professor of communication studies and computer science at Northwestern University, told me. “I’ve caught myself starting to read the generative results, and then I stop myself halfway through. I’m like, Wait, Nick. You can’t trust this.”

Nayak said the team focuses on the bigger underlying problem, and whether its algorithm can be trained to address it.

If Nayak is right, and people do still follow links even when presented with a snippet, anyone who wants to gain clicks or money through search has an incentive to capitalize on that—perhaps even by flooding the zone with AI-written content.

Nayak told me that Google plans to fight AI-generated spam as aggressively as it fights regular spam, and claimed that the company keeps about 99 percent of spam out of search results.

The result is a world that feels more confused, not less, as a result of new technology.

The Kenya result still pops up on Google, despite viral posts about it. This is a strategic choice, not an error. If a snippet violates Google policy (for example, if it includes hate speech) the company manually intervenes and suppresses it, Nayak said. However, if the snippet is untrue but doesn’t violate any policy or cause harm, the company will not intervene.

experts I spoke with had several ideas for how tech companies might mitigate the potential harms of relying on AI in search

For starters, tech companies could become more transparent about generative AI. Diakopoulos suggested that they could publish information about the quality of facts provided when people ask questions about important topics

They can use a coding technique known as “retrieval-augmented generation,” or RAG, which instructs the bot to cross-check its answer with what is published elsewhere, essentially helping it self-fact-check. (A spokesperson for Google said the company uses similar techniques to improve its output.) They could open up their tools to researchers to stress-test it. Or they could add more human oversight to their outputs, maybe investing in fact-checking efforts.

Fact-checking, however, is a fraught proposition. In January, Google’s parent company, Alphabet, laid off roughly 6 percent of its workers, and last month, the company cut at least 40 jobs in its Google News division. This is the team that, in the past, has worked with professional fact-checking organizations to add fact-checks into search results

Alex Heath, at The Verge, reported that top leaders were among those laid off, and Google declined to give me more information. It certainly suggests that Google is not investing more in its fact-checking partnerships as it builds its generative-AI tool.

Nayak acknowledged how daunting a task human-based fact-checking is for a platform of Google’s extraordinary scale. Fifteen percent of daily searches are ones the search engine hasn’t seen before, Nayak told me. “With this kind of scale and this kind of novelty, there’s no sense in which we can manually curate results.”

Creating an infinite, largely automated, and still accurate encyclopedia seems impossible. And yet that seems to be the strategic direction Google is taking.

A representative for Google told me that this was an example of a “false premise” search, a type that is known to trip up the algorithm. If she were trying to date me, she argued, she wouldn’t just stop at the AI-generated response given by the search engine, but would click the link to fact-check it.