Artificial Intelligence: Last Week Tonight with John Oliver (HBO)
Artificial Intelligence: Last Week Tonight with John Oliver (HBO)
Artificial intelligence is increasingly becoming part of our lives, from self-driving cars to ChatGPT. John Oliver discusses how AI works, where it might be heading next, and, of course, why it hates the bus.
Connect with Last Week Tonight online…
Subscribe to the Last Week Tonight YouTube channel for more almost news as it almost happens: www.youtube.com/lastweektonight
Find Last Week Tonight on Facebook like your mom would: www.facebook.com/lastweektonight
Follow us on Twitter for news about jokes and jokes about news: www.twitter.com/lastweektonight
Visit our official site for all that other stuff at once: www.hbo.com/lastweektonight
Almost a decade ago, I attended a job fair which had a resume specialist. The subject of the seminar was improving chances of your resume being noticed by employers. The specialist’s only real advice was to cut and paste the entire job listing that you were applying for into your resume in either a header or footer, change the text to white, and reduce it to one point font size. This way, the algorithms that scan each resume would put yours at the top of the list because it had all the keywords it was programmed to find.
@Just Anomanwould you recommend the death penalty? What additional penalties would you recommend? They’re already being passed over for the job.
That’s basically how old school search engines worked until GOOGLE switched to referential rankings.
honestly brilliant
George Santos has a resume company if you want to use it
@plushnyt Oh brother.
The Tay AI also made the funniest tweet ever. She said that Ted Cruz wasn’t the zodiac killer cause Ted Cruz would never be satisfied with the deaths of only 5 innocent people
@Grigori Belov “I probably like Ted Cruz more than any of my colleagues, and I hate Ted Cruz.”
@Grigori Belov Tay got that right.
She does not like his stupid chin
Totally correct.
Holy shit
A note, less on the subject matter and more on John’s delivery of the lines … I really admire how he can say “Sometimes I just let this horse write our scripts, luckily half the time you can’t even tell the oats oats give me oats yum” without skipping a beat or losing face. Now THAT’S professionalism.
I’m disappointed that that joke didn’t get brought back later.
I was at the taping of this episode. It’s surprising how few takes he needs to get through an episode. He nailed that line without breaking a sweat.
thanks for the translation, I thought he was just making random funny noises
@StilasCzech I think I was going for “losing his facial expression”, but yes, this is pretty on point too.
Just like Ron Burgundy, John will read absolutely ANYTHING you put on that teleprompter
TRUE STORY: In my teens wanted to work at a movie theater – and they handed applicants a mind-numbing 14 pages application – wanting to know everything about you – even what hobbies and sports you liked – it was entirely ridiculous – around page 8, I got worn out from filling out this ‘essay’ of my life for a stupid theater job – SO when I got to the section asking if I had ever been arrested before = I said: “Yes, I murdered an entire movie theater crew for asking way too many questions, but got off on a technicality.” – and turned that resume into the manager as I stormed out the door, pissed off that I had wasted an hour of my time filing out paperwork w/o an interview.
…well,
2 days later I got a call to come back to the theater for an interview, and thought, oh sh*t, well, I guess I’m going to get railroaded and berated by the management for my saucy comment – but I showed up anyways so that at least I could suggest that they TONE DOWN the length of their stupid applications.
…turns out, they offered me a job, so I asked the most obvious question:
“So, you read my application … all of it?”
“Oh yes, looks good” the manager responded
and I knew they were a bunch of lying dimwits ~ I ended up working there for the next 5 yrs, and eventually rose in ranks to become the theater manager –
When I told my story to new recruits that nobody reads the stupid applications – they scoffed and didn’t believe me – so I took them to the locked office storage and rifled through the stuffed cabinets of folders of all the applications they kept and found mine, and showed it to them to their amazement.
Applications are a farce, you get hired by chance and immediate need.
…
I always thought that if I every flipped out and murdered my entire staff, at least I could say that I didn’t lie on my resume.
Maybe it was some kind of weird data mining
The next time I want you to make my popcorn.
You are the Lenny Bruce of the workplace. A genius.
Wow what a great story !!!!
@Samfia Drangus iunno man I managed a shipping yard, and come to find out they never checked a single resume, never officiated a single reference, and I kinda always knew, single you can’t put immediate family I would always as a teen put my uncle’s or deacons from my grandpa’s church, and I would occasionally ask if anyone called about my references….not once ever. And I started in the work force at 16 in 2002, I can nearly guarantee 95% of any job, they ain’t checking any of that crap. And it’s stupid and lessens my faith in humanity to put someone thru that kind of crap for no real reason.
I love how John said “our” scripts. He’s very aware that he’s a face and and mouth peace for a whole team of talented people.
Whoever on your staff came up with the animation of Clippy deserves a raise.
@Jon Paul Cuellar for me it was the end of the wire heating up due to friction HAHAHA
And perhaps a shoulder to cry on
If only to pay for their desperately needed therapy.
being a part of some show’s in-house shitpost graphics team is my dream job ngl
Well they did get a raise out of Clippy
i’m so glad he touched on the significant issue of people observing ai as “unbiased” simply bc it’s not human. where do they think the data came from?
@Scipio Africanus I think you misunderstood. He didn’t say people don’t have bias. What he said is that people can detect biases that don’t align with their own biases.
There is something to be said for humans making claims of bias when there is none. AI has the potential to help us get outside our illogical bubbles of “mexicans are stealing our jobs” or “men are sexist if they don’t like feminism”.
And as for putting more and more layers of filters on chatgpt to make it more sanitized, I can give you a personal example: I asked it to write a story about an alien coming to earth, and at one point I asked it to write about it falling in love with a beautiful girl from Russia. Over and over again it screamed at me that “this is inappropriate” and “this is harmful language” and I could not figure out what was wrong with my prompt. I changed one word, and it was happy. That word was “beautiful”. Imagine thinking that a proper AI should not be allowed to describe a girl as beautiful?
Garbage in. Garbage out.
@Robert Beenen “people… don’t have … bias” Sorry. Your sentence does not compute.
Of course that only works when the people observing the system don’t have the same bias.
That whole thing with the AI learning that rulers were an indication of malignant skin cancer reminds me of when the Russians trained dogs to crawl under tanks with explosives to blow them up, but since they trained them on their own tanks they didn’t run to German tanks, but instead ran to Russian ones.
I watched someone literally cyber-bully an instance of Chat GPT into saying 2 + 2 =5. These bots are interesting and potentially useful but very easily abused or misused in some pretty scary ways.
Yes saw that too, and interestingly, the way bing seems to protect against this kind of manipulation is scary in its own way, it becomes *really* agressive when you disagree with its answer.
One of my favorite sci-fi webcomics once described AI as ‘a force-multiplier for stupidity’. That is, the fact that they tend to be really good at a narrow scope selection of things, which rarely included ‘critical thought’ or ‘questioning authority’, means that they can be used by incredibly dumb people to cause a *lot* more damage than they’d otherwise be able to. (Which is quite a lot to begin with, as I’m sure we’ve all learned first-hand.)
@Arnaud Installe Unfortunately AI’s, at least as we currently think about them, are not that difficult to create now that the hard research is done. I fear that regulations won’t ultimately matter for AI, since all the data is out there and processing only gets ever cheaper, so that anyone can train their own models for whatever purposes.
Midjourney might generate a getty images watermark which proves the use of getty’s images, but take a guess whether the next dataset will have watermarks in them. How do you judge copyright in such a case. And how do you enforce whatever laws they come up to AI?
Nevermind watermarks, it only takes one fool of a person to make a virus-writing AI and unleashes it to internet with no safeguards to cause incredible damage.
Indeed. That’s why you shouldn’t have AI models created by anyone, but by people who hold a PhD in AI.
Those people often being incredibly POWERFUL dumb people