Should we worry about AI and algorithms in government? | Lyria Bennett Moses | TEDxSydneySalon

As confidence in — and the proliferation of — machine learning technologies, algorithms and AI grows, governments around the world are increasingly relying on them to make decisions. Decisions about how to allocate resources, in the justice system and of course, in surveillance. But there is a long list of examples where this technology has raised serious concerns. So, what is needed before we trust government to use these tools when making decisions that affect us? Lyria Bennett Moses argues that what we need isn’t more ‘ethical AI’ statements or even technological improvements — instead it’s about legal constraints built into legislation, protection for privacy and autonomy, detailed standards that explain ‘how’ to get it right, and engagement with an educated citizenry.

[clusterstage]

3:06 yes, in fact, its not the AI we should be scared of, but the people in AUTHORITY who listen to the recommendations of this algorithm, without even questioning how it arrived at that conclusion. There are too many cases to work with and hence the laziness of algo-verification.

[raulguilhermealves]

yes, we should already be freaking out. But we're not. And that's the problem

[teebone]

Even if we have a social credit system in the United States you know billionaires will just bribe politicians to keep themselves out of it

[originaljungle]

This is John Connor. If you’re listening to this, you are the Resistance

[duke]

YES!! WE should afraid, VERY afraid! " Intelligence with out COMMON SENSE is very dangerous " I don't think that the people who are creating this AI can ever put the "genie" back in the bottle.

[clusterstage]

Historical data 4:17 isn't even close to 90% accurate in predicting the future.
The AI researchers who designed that algorithm may have overlooked at the "gamblers fallacy", which is a backwards way of relying on the past to predict what's gonna happen tomorrow.

[clusterstage]

in about 11 minutes of your talk, you mentioned about laws. if you think about it, laws can't quite catch up to slow developing tech like the metaverse. By the time they restrict AI with the 1st approved law, AI will have evolved to be almost unrecognizable.
Even to the point the law MAY NO LONGER apply to that year.

Clearly you're unaware why politics is irrelevant to the speed of AI.
Ironically, I love how your talk made me realize a few things myself.

[ollydsouza]

Risk Assessment? It relies on "feedback"! Decision making is done within an operational context! And this process workflow has to be simulated and checked out before release!

[deamdeam]

I am sure if in the US all court systems would be replaced by the AI, every person in our corrupt government would be already in jail and most of the people in jail be out of prisons.

[KatinaMichael]

Such an important, informative and well thought out TEDx talk! Congrats Lyria Bennett Moses. I'll be adding it to my course outlines.

[China-]

“To save the world requires faith and courage: faith in reason, and courage to proclaim what reason shows to be true.”
*― Bertrand Russell🇬🇧*

[ollydsouza]

I think we are missing the point here - we need closed loop processes to drive any operation, self correcting, advisory tools as part of the WORKFLOW!

[clusterstage]

12:35 unfortunately, code being "released transparently" doesn't mean that the algorithm will allow changes in its codebase.
Plus, no one has attention span to pay EVEN a closer look on reviewing the code.

I've designed a video to test our attention span.
It’s sooo boring that the higher we score, the lesser damage done to us by the algorithm in the apps we use.
Try this; comment the time you quit watching and let's see the results.

[jasonking]

We should worry about government in government.

[شهرزاد_نور_محمدي]

My father had brought me here (at GKB Graveyard) just before I had left Pakistan to enter Harvard University in 1969 (and said):

"Remember, whatever happens to you, you will ultimately return here. Your place is here. Your roots are here. The dust and mud and heat of Larkana are in your bones. And it is here that you will be buried!"
*—Benazir Bhutto 🇵🇰*

[AnthatiKhasim-ie]

"SmythOS supports a wide range of applications, ensuring compatibility with your favorite software, whether for work, gaming, or creative projects."

[EPY]

Haven't watched it yet: YES. Let's see if she changes my mind

[clusterstage]

That's why they invented AI "recommendations"; it's easier to direct a person's attention than predict a person's behavior.

[China-]

“I would never die for my beliefs because I might be wrong.”
*— Bertrand Russell🇬🇧*

[hmstr]

Technology is our friend as long as we use it responsibly.