Almost Timely News: Is AI Taking Jobs? Maybe (2023-07-16)
Almost Timely News: Is AI Taking Jobs? Maybe (2023-07-16) ::?View in Browser
Content Authenticity Statement
95% of this newsletter was generated by me, the human. You will see some AI artwork in the opening section.
Watch This Newsletter On YouTube ??
What’s On My Mind: Is AI Taking Jobs? Maybe
I think I’ve answered the “Will AI take your job?” question more times in the past week than in the past 6 months. Part of that was because I was on a bunch of podcasts, and part of that was the headlines, the news stories of the week. The Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA) went on strike this week. One of the points of contention for SAG-AFTRA - that they share with the Writers’ Guild of America - is their concerns about whether AI will take their jobs.
So let’s have a nuanced discussion about the topic, because it isn’t as clear-cut as many folks would like it to be. I talked to one person who was steadfast that we should only remain positive about AI, and another person who believes it to be an existential threat to humanity. The truth is somewhere in between.
AI, as it stands today in mid-2023, isn’t going to be taking anyone’s job. Even the most sophisticated models and ensembles cannot do the disparate parts of most jobs. AI is quite capable of doing specific tasks, and doing those tasks well. Those capabilities increase every day; this week,?Anthropic released its Claude-2 model ?which is a huge leap forward for large language models. With a 100,000 token context window-
Okay, probably should explain that part. A context window is effectively how much a large language model can remember at any given time in the context of a specific conversation. If you’ve used smaller models like GPT-J-6B, the old GPT-3, and many of the open source models, you know that these models have relatively short memories. You can be talking to them about something and in the span of a few paragraphs, it’s like they’ve forgotten what they were talking about. That context window is the working memory. Older models and open source models have context windows of about 2,000 tokens. Tokens are word fragments; on average, 100 tokens equals about 66 words, or about 2/3 the token size. A model that has a context window of 2,000 tokens can remember about 1,300 words, give or take.
That means if you get into a lengthy conversation that’s more than a page of text, the model starts to lose its memory. Details you discussed previously it will suddenly not know. And that can be really frustrating if you’re working with documents longer than a page.
The current OpenAI models in ChatGPT support context windows of up to 16,000 tokens for GPT-3.5-Turbo (which is the default model for ChatGPT) or up to 32,000 tokens for GPT-4 in the paid version of ChatGPT. If you do the math, that means ChatGPT and software based on its underlying models can handle about 10,000 words at a time for the default model and around 20,000 words at a time for the paid version. That makes for more satisfying conversations, more capabilities, longer content creation, the works. Bigger context windows, all other things being equal, tend to be better.
So when Anthropic released its GPT-4 competitor, Claude 2, with a 100,000 token context window - equal to about 66,000 words - that was a very big deal. You could feed it an entire business book or fiction work as a prompt, for example, and tell the model to rewrite the entire book in the style of Ernest Hemingway.
What does this all have to do with your job? The bigger and more capable models get, the more tasks they can handle. Every time we have a big leap forward in model capabilities, that opens the door for us to hand off more tasks to AI. Does your book draft need a sensitivity reader or a first-pass editor? Feed it to a model with a suitably large context window and have it do the initial work. Do you want to rewrite a work of fiction you wrote in one universe to another universe? The largest models can handle that task. Do you want to write thousands of lines of code? Also doable. In fact,?GPT-4’s Code Interpreter, which I wrote about earlier this week, is absolutely mind-melting in how good it is .
What we - and by we, I mean most AI practitioners - have been saying for quite some time now is that AI isn’t going to take your job, but a person skilled with AI will take the job of a person who isn’t skilled with AI. That’s… sort of true. Again, there’s nuance. There are some jobs, some content creation jobs, where AI will absolutely take that job if it’s valuable enough to do so. This week, SAG-AFTRA reported that the Alliance of Motion Picture and Television Producers (AMPTP)?allegedly included in their negotiating points , this:
“This ‘groundbreaking’ AI proposal that they gave us yesterday, they proposed that our background performers should be able to be scanned, get one day’s pay, and their companies should own that scan, their image, their likeness and should be able to use it for the rest of eternity on any project they want, with no consent and no compensation. So if you think that’s a groundbreaking proposal, I suggest you think again.”?- Duncan Crabtree-Ireland, chief negotiator for SAG-AFTRA
Now, no one seems to be able to produce the actual document where this is written, but the perspective alone is worth considering. Yes, with today’s technology, it is possible to scan a person’s likeness and re-use that person in perpetuity. I should hope anyone in the entertainment industry has a good enough lawyer to look for that clause in a contract and amend it appropriately. But for background talent, our technology is getting good enough that background actors (also known as extras) can be largely synthetic anyway. That job - a person milling around in the background - is one that AI absolutely can do. If you haven’t already seen?Unreal Engine’s Metahuman Creator (here’s a short film made entirely with the tech) , you should. It’s uncanny how good the generated humans look - more than good enough to synthesize a background actor wandering down a street or standing on a corner looking at their phone.
So yes, there are some instances where AI will take someone’s job. Let’s now talk about the second part, the idea that someone skilled with AI will take the job of someone who is not. This is true, but there’s an additional dimension at play here.
AI is a force multiplier. It lets you be more of you, it amplifies your human capabilities. A good writer, with the help of AI, becomes a prolific good writer. A good painter, with the help of AI, becomes a prolific good painter. AI works best when someone who has subject matter expertise can craft the necessary prompt details to bring out the specifics that only an expert would know. For everyone else, it adds to our capabilities, gives us capabilities that we don’t have.
For example, in a recent Discord chat, some friends of mine were celebrating two members of our community becoming a couple. One of them has an avatar of a blue frog. The other has an avatar of a trash can (don’t ask). In the conversation, someone joked that they needed a combined profile picture of some kind. Naturally, I hopped over to Bing Image Creator and gave it the appropriate prompt to generate:
AI gave me a capability I don’t have. I’m not an artist. I don’t generate art like this. The software, however, enabled me to become a good enough artist to fulfill the requirements in that moment. Is it great art? No.?Could a human artist, a skilled artist, have done better? Yes.
Is it good enough? Yes.
AI is a force multiplier. Which in turn means it allows one person to do the work of more than one person. A writer, empowered with AI, can do the work of more than one writer who doesn’t have AI capabilities. How much more? It depends, but it’s not unreasonable to believe that it’s multiples - 2, 3, 5, maybe even 10 people. Which means if you’re, say, a content marketing production agency or company, you could either scale your business 2, 3, 5, or 10x if there’s enough business to be had, or alternately reduce headcount by 2, 3, 5, or even 10x depending on the content you create and how skilled your writers are.
This is the part we’re not totally being honest about when we say a person skilled with AI will take the job of a person not skilled with AI. It’s not a one-to-one ratio. Depending on the job, it could be a many-to-one ratio.
Now, is it all doom and gloom? No.?For every job AI consumes, it will create ripple effects,?which we’ve talked about in the past . You might lose 90% of your writers but then you have to hire 10x your number of editor, or promoters, or distributors, etc. A massive change in efficiency in one part of your supply chain will have upstream and downstream effects on the rest of the supply chain.
But there will be impacts that are greater than the optimists are predicting, and lesser than the nihilists are predicting.
So what? What’s the antidote, what’s the strategy, what’s the play to keep you safe? It’s what we’ve been saying all along - the person skilled with AI takes the jobs of people not skilled with AI. Right now, things are still in flux. The market isn’t settled yet. There isn’t a calcified hegemony in place with permanent winners and losers. That means there’s still time for you to carve out your niche, as an AI-empowered worker no matter what industry you’re in. That window is closing, but you still have time to skill up, to learn, to explore, and to be a leader in your space.
The AMPTP may not hire background actors in the future, but they will absolutely hire someone skilled at Unreal Engine to build metahuman background talent for productions. You want to be that person.
Got a Question? Hit Reply
I do actually read the replies.
Share With a Friend or Colleague
If you enjoy this newsletter and want to share it with a friend/colleague, please do. Send this URL to your friend/colleague:
ICYMI: In Case You Missed it
Besides the newly-refreshed?Google Analytics 4 course I’m relentlessly promoting (sorry not sorry) , I recommend the livestream I did with Katie and my Save Warrior Nun colleagues this week on how data and analytics helped Save Warrior Nun.
I also started a new column, thanks to AI capabilities, called Borrowed Time, in which I showcase and thank the people who have had me on their shows. Be sure to check those out as well.
Skill Up With Classes
These are just a few of the classes I have available over at the Trust Insights website that you can take.
Premium
Free
领英推荐
Get Back to Work
Folks who post jobs in the free?Analytics for Marketers Slack community ?may have those jobs shared here, too. If you’re looking for work, check out these recent open positions, and check out the Slack group for the comprehensive list.
Advertisement: Bring My AI Talk To Your Company
I’ve been lecturing a lot on large language models and generative AI (think ChatGPT) lately, and inevitably, there’s far more material than time permits at a regular conference keynote. There’s a lot more value to be unlocked - and that value can be unlocked by bringing me in to speak at your company. In a customized version of my AI keynote talk, delivered either in-person or virtually, we’ll cover all the high points of the talk, but specific to your industry, and critically, offer a ton of time to answer your specific questions that you might not feel comfortable asking in a public forum.
Here’s what one participant said after a working session at one of the world’s biggest consulting firms:
“No kidding, this was the best hour of learning or knowledge-sharing I’ve had in my years at the Firm. Chris’ expertise and context-setting was super-thought provoking and perfectly delivered. I was side-slacking teammates throughout the session to share insights and ideas. Very energizing and highly practical! Thanks so much for putting it together!”
Pricing begins at US$7,500 and will vary significantly based on whether it’s in person or not, and how much time you need to get the most value from the experience.
What I’m Reading: Your Stuff
Let’s look at the most interesting content from around the web on topics you care about, some of which you might have even written.
Social Media Marketing
Media and Content
SEO, Google, and Paid Media
Advertisement: Business Cameos
If you’re familiar with the Cameo system - where people hire well-known folks for short video clips - then you’ll totally get Thinkers One. Created by my friend Mitch Joel, Thinkers One lets you connect with the biggest thinkers for short videos on topics you care about. I’ve got a whole slew of Thinkers One Cameo-style topics for video clips you can use at internal company meetings, events, or even just for yourself. Want me to tell your boss that you need to be paying attention to generative AI right now?
Tools, Machine Learning, and AI
Analytics, Stats, and Data Science
All Things IBM
Dealer’s Choice : Random Stuff
Advertisement: Ukraine ???? Humanitarian Fund
The war to free Ukraine continues. If you’d like to support humanitarian efforts in Ukraine, the Ukrainian government has set up a special portal, United24, to help make contributing easy. The effort to free Ukraine from Russia’s illegal invasion needs our ongoing support.
How to Stay in Touch
Let’s make sure we’re connected in the places it suits you best. Here’s where you can find different content:
Events I’ll Be At
Here’s where I’m speaking and attending. Say hi if you’re at an event also:
Events marked with a physical location may become virtual if conditions and safety warrant it.
If you’re an event organizer, let me help your event shine.?Visit my speaking page for more details.
Can’t be at an event? Stop by my private Slack group instead,?Analytics for Marketers .
Required Disclosures
Events with links have purchased sponsorships in this newsletter and as a result, I receive direct financial compensation for promoting them.
Advertisements in this newsletter have paid to be promoted, and as a result, I receive direct financial compensation for promoting them.
My company, Trust Insights, maintains business partnerships with companies including, but not limited to, IBM, Cisco Systems, Amazon, Talkwalker, MarketingProfs, MarketMuse, Agorapulse, Hubspot, Informa, Demandbase, The Marketing AI Institute, and others. ?While links shared from partners are not explicit endorsements, nor do they directly financially benefit Trust Insights, a commercial relationship exists for which Trust Insights may receive indirect financial benefit, and thus I may receive indirect financial benefit from them as well.
Thank You
Thanks for subscribing and reading this far. I appreciate it. As always, thank you for your support, your attention, and your kindness.
See you next week,
Christopher S. Penn