...One question - has the increase in total DL outpaced the increase in size of your port - either in comparison to previous port size or compared to Adobe as a whole?...

My portfolio isn't that large (2,335 today) and hasn't grown much in the last couple of years. I haven't kept track of portfolio growth by date as it's been so (a) erratic - I do this part time as I have time to spare, and (b) small.

Bottom line is that my growth in sales isn't because I've been an upload demon And without knowing exactly what Adobe's growth has been, I have been keeping track of some stats (started when the free collection started) and April 2022 there were 190,203,517 photos and May 2023, 202,460,892 (I have overall size, but that included Pond5 videos going away which skewed things a bit). That was about 6.5% growth and is much less than the sales growth I'm seeing.

I did make PNGs of some of my isolated photos last year sometime which is sort of growth but sort of not (as in the images didn't change, just the availability in a useful format), so you could subtract about 50 from my port size for that

This is only available in the beta version of Photoshop (and I have no idea how one gets that):

https://techcrunch.com/2023/05/23/adobe-brings-fireflys-generative-ai-to-photoshop/

"Photoshop is getting an infusion of generative AI today with the addition of a number of Firefly-based features that will allow users to extend images beyond their borders with Firefly-generated backgrounds, use generative AI to add objects to images and use a new generative fill feature to remove objects with far more precision than the previously available content-aware fill."

https://www.theverge.com/2023/5/23/23734027/adobe-photoshop-generative-fill-ai-image-generator-firefly

According to Gizmodo, "Adobe�s Firefly Image Generator Is Going to Make Photoshop Much Easier to Use. Soon, even your grandparents could be Photoshop experts"

https://gizmodo.com/adobe-firefly-ai-image-generator-photoshop-generative-f-1850462988

https://arstechnica.com/information-technology/2023/05/adobe-photoshops-new-generative-fill-ai-tool-lets-you-manipulate-photos-with-text/

https://www.businessinsider.com/sc/adobe-photoshop-unlocks-a-new-era-of-generative-creativity-with-firefly

https://9to5mac.com/2023/05/23/photoshop-first-adobe-app-with-generative-ai/

https://www.digitalcameraworld.com/news/adobe-integrates-fireflys-generative-ai-with-photoshop

Still nothing more about the compensation model for the Adobe Stock contributors without whom none of this would be in beta...

"Through the GIPHY acquisition, we are extending our audience touch points beyond primarily professional marketing and advertising use cases and expanding into casual conversations"

https://investor.shutterstock.com/news-releases/news-release-details/shutterstock-acquire-giphy-worlds-largest-gif-library-and-search

https://investor.shutterstock.com/static-files/2899adce-d747-4cc5-964f-8a2134df94c7

https://finance.yahoo.com/news/meta-sells-giphy-53m-shutterstock-165241138.html

The blurbs include mention of Sponsored GIFs - which I assume means that while end users won't pay, companies will pay SS to create custom (brand-focused) GIFs, hoping for a viral moment.

If you can figure out what this word salad means, then you know how SS will make its $53 million back:

"We expect to execute against multiple monetization opportunities across advertising, content and distribution"

Some comments

https://techcrunch.com/2023/05/23/meta-sells-giphy-to-shutterstock-for-53m-after-uk-divestment-order/

WSJ (paywall) "Shutterstock said it plans to fund the deal, slated to close in June, with cash on hand and its revolving credit facility, adding that Meta is entering into an API agreement to ensure continued access to Giphy�s content across its platform."

https://www.wsj.com/articles/facebook-parent-meta-agrees-to-sell-giphy-for-53-million-on-regulators-order-7b1c2ec6

https://www.theverge.com/2023/5/23/23734220/shutterstock-giphy-meta-cma-uk-antitrust

From The Verege article: "For Shutterstock, the deal will augment its content library, expanding it to include GIFs and Stickers. The company also said it would help its �generative AI and metadata strategy.� That�s a bit more vague, but Shutterstock is currently using generative AI to provide imagery to its customers. Acquiring Giphy�s substantial library could theoretically give Shutterstock reams of training data for future AI-based products."

From the Variety story "This is an exciting next step in Shutterstock�s journey as an end-to-end creative platform..."
https://variety.com/2023/digital/news/giphy-sold-meta-shutterstock-1235622548/

From Engadget: " "This may affect Shutterstock's customers, though. Chief executive Paul Hennessy hopes Giphy will help commercialize Shutterstock's GIF collection � don't be surprised if animated images play a prominent role in Shutterstock's offerings."

https://www.engadget.com/metas-paid-verification-system-comes-to-the-uk-162523922.html

 This is a legal decision on what level of transformation constitutes fair use. As fair use is one of the arguments offered as to why generative AI images trained on copyrighted work is legally OK, I think it's worth adding to this list

https://petapixel.com/2023/05/18/supreme-court-rules-andy-warhols-prince-art-is-copyright-infringement/
https://www.hollywoodreporter.com/business/business-news/supreme-court-andy-warhol-prince-copyright-1235495647/
https://www.billboard.com/pro/andy-warhol-prince-supreme-court-copyright-case-ruling/
https://www.cooley.com/news/insight/2023/2023-05-18-supreme-court-rules-andy-warhols-prince-portraits-not-fair-use
https://www.theguardian.com/artanddesign/2023/may/18/andy-warhol-copyright-prince-paintings-lawsuit
https://variety.com/2023/music/news/andy-warhol-prince-images-supreme-court-copyright-1235617622/

Shutterstock held a virtual "showcase" Wednesday afternoon and the video is up on YouTube

https://www.youtube.com/watch?v=RbZ5q3X6MSY

I watched some of it (and skipped some) and noted a couple of interesting things.

Skip Wilson hosted (VP Brand Marketing, formerly at Peloton).

About 2 minutes in, a panel discussion led by the chief product officer ostensibly about creative possibilities with AI. Ms. Schoen layered the buzzwords - disruption, innovation, etc. - and said "...generative technology is magical..." Please... There really wasn't much real content - panel members were from Peloton, NVIDIA and AWS.

Perhaps Skip Wilson's prior gig was why Jackie Moore, a producer from Peloton was there. She said she couldn't talk about the specifics but they were using AI in their production (so why have her on the panel??). I listened for a bit but then skipped to the next segment.

About 30 minutes in, Skip Wilson announces a partnership with AI for Good - responsible AI, blah, blah...

https://www.prnewswire.com/news-releases/shutterstock-and-itus-ai-for-good-collaborate-to-advance-responsible-ai-301826531.html

At about 33 minutes, an interview with Joh Reynolds (artist, https://twitter.com/ArtByJah) about his use of AI

About 43 mins is a product announcement/demo of an AI design assistant which looks pretty interesting. The slightly over-eager chat from the bot could get old fast, but it looks like it might be useful to customers with no design staff.

At 51 minuets there's a final panel discussion.

They emphasize their ethical approach to AI and talk about the contributor fund and how Shutterstock "helps" artists make money from their work. There was a discussion of the process of offering customers indemnity for the work they generate and then license through SS. It appears they'll run the output through their inspection process and if the content passes, the customer will have legal indemnity from SS in the event there's a lawsuit.

The discussion contrasted how SS's approach to generative AI was different in that they're doing it ethically and compensating contributors (no mention that they did the initial training before telling anyone about it). That's certainly Adobe's pitch too. Based on my experience with Firefly beta, Shutterstock will clean Adobe's clock with their design tool. The demo was undoubtedly showing off a carefully scripted and tested scenario, but Firefly is an exercise it time consuming frustration trying to get anything usable that's even close to what the prompts asked for.

You can join their waitlist to try the creative AI tool

https://www.shutterstock.com/ai-image-generator

I have a 2004 Shutterstock model release, and the form is essentially the same. Other agencies have the same idea, but it's expressed differently and thus sounds less worrying.

I believe the reason this hasn't caused any problems is that a model cannot grant rights to anyone that the model doesn't have, and thus there is no way they can grant SS copyright in a photograph when the model doesn't have that in the first place.

The essential part of that sentence is that the model is not making a claim on the copyright; agreeing that the rights are "...free and clear from any claims by me or anyone acting on my behalf."

I decided this should be its own topic

https://www.microstockgroup.com/fotolia-com/googleadobe-partnership-with-bard-firefly/

Shutterstock had a mini-rally earlier this week, but closed today at $52.63 (down $2.11 from Wed)

https://finance.yahoo.com/news/peri-sstk-better-value-stock-154003633.html

For some reason I can't fathom - other than the investor newsletter just has no clue about the businesses these companies are actually engaged in and how they work (who the customers are) - this article put Shutterstock and Perion in the same business category:

https://finance.yahoo.com/news/peri-sstk-better-value-stock-154003633.html

I looked at Perion's home page and it's full of buzzy fluff about capturing digital dollars, or fragmenting technology (slight misrepresentation on my part):

"Perion is focused on the future, and has built a tech-stack that drives our Capture & Convince business solutions. Our AI and machine learning can be applied to sourcing and optimizing traffic, transforming dumb funnels into smart consumer journeys."

I don't think they're in the same business at all, but at the end, the article about taking advantage of undervalued stocks recommends buying Perion's stock, not SSTK

"Both PERI and SSTK are impressive stocks with solid earnings outlooks, but based on these valuation figures, we feel that PERI is the superior value option right now."

Here's a more general article about what Perion does - I have no love for SSTK, but it must be frustrating to have investment advisers lump you in with the wrong bucket and then drag the stock down as a result. But them's the breaks when you decide to go public

https://seekingalpha.com/article/4584594-perion-network-still-very-discounted


https://www.perion.com/

Edited to add that the stock closed down a bit more Tuesday 16 May - $51.29, (down $1.67 from Monday).

Tomorrow afternoon SS is hosting a (streaming) Gen AI event "AI is the game changer of all game changers�something every brand and business in the industry can gain from. Don�t forget to book your seat so you can learn how to make it work for you." They're clearly hoping if they bang the AI drum enough it'll do them some good.

Separately I saw a marketing piece - misspelled drivel, IMO - that says there's huge growth ahead. Probably not all that well researched as they include (the now bankrupt) EyeEm in the list of "top companies"

https://www.digitaljournal.com/pr/news/a2z-market-research/stock-photos-and-videos-market-to-witness-huge-growth-by-2029-adobe-stock-shutterstock-fotosearch

"Stock Photos and Videos Market is growing at a +7% CAGR during the forecast period 2023-2030. The increasing interest of the individuals in this industry is that the major reason for the expansion of this market"

Edited May 26 to add that SSTK closed yesterday at $48.94, so there's clearly something bothering investors. Their "ex-dividend date" (last day to be qualified to receive the next stock dividend) is May 31 and typically that causes the stock price to go up a bit as people buy...

May 30 close, $47.22

...Because of Adobe Firefly i would invest in Adobe and not in Shutterstock.

Have you used the Firefly beta?

IMO it is unusable for most people for most purposes at the moment.

This Wall Street Journal article (paywall) is about AI and advertising - along with the turmoil resulting from the shift from traditional TV to streaming and how the advertising dollars are following (or not). The reason I think this might have an impact on stock agencies is that a good chunk of our licenses are for advertising uses.

Digital Video Publishers Tout AI and New Metrics at NewFronts Sales Events
https://www.wsj.com/articles/digital-video-publishers-tout-ai-and-new-metrics-at-newfronts-sales-events-97a42c4f

A couple of interesting quotes:

"Players in free ad-supported television or ad-supported video on demand, smart-TV manufacturers like LG and Samsung, premium video publishers and others at the NewFronts all reported growing viewership and sought to convince buyers they deserved video ad spending..."


"As Writers Guild members marched outside of Peacock�s presentation, in part protesting studios� reluctance to regulate AI-generated material in television and movies, media executives sold AI as a technology to elevate buyers� advertising deals.
Roku described a new AI-driven feature that scans programming on its ad-supported Roku Channel to match advertisers, their products and their campaigns with relevant scenes in shows and movies.
�Imagine promoting a new phone line in the ad break after every time E.T. phones home, or advertising an apparel line every time Tim Gunn says �make it work� on �Project Runway,�� Julian Mintz, Roku�s head of U.S. brand sales, told ad buyers.""

The current head of the US Federal Trade Commission wrote an op-ed piece for the New York Times (paywall) but there's a good summary in Ars Technica

https://arstechnica.com/information-technology/2023/05/we-must-regulate-ai-ftc-chair-khan-says/

There aren't very many comments in the NY Times - all comments are moderated, but even so only 267? An article about Tucker Carlson's texts had 2617 Comments are pretty negative about her approach - either because they are opposed to regulation in principle or because they don't see anything sufficiently specific about how to rein in tech companies. An example:

"There isn't a single specific suggestion in this article about how to regulate the use of AI.  It just lists the problems with which we are all already familiar.  If the head of a federal agency is going to place an article in the NYT, it should offer more than this."

There was one comment about content creators: "not a peep in this opinion about protection of intellectual property and creators work.   Just like streaming music, no doubt the AI lobbyists are on the job right now, making sure govt looks the other way as their clients appropriate the value of creative work."

I'm beginning to see some common themes in the arguments for or against being able to copyright AI-created works, who is the creator of a work, and various analogies to past technologies and whether they produced original work.

Here's another article on the subject, again behind a paywall, arguing that the US copyright office's decision in the graphic novel was wrong. I don't agree with the point of view - I'm still stuck on the fact that none of the current generative AI tools would be able to produce if they hadn't ripped off masses of work to "train" themselves. The comments are lengthy and in many ways more interesting that the opinion piece The author also wrote �Creators Take Control: How NFTs Revolutionize Art, Business, and Entertainment.�. I haven't read that, but it doesn't surprise me that these two points of view come from the same person.

https://www.washingtonpost.com/opinions/2023/04/27/artificial-intelligence-copyright-decision-misguided/

In talking about the use of prompts as creative intellectual labor, he writes:

"If the creator doesn�t like the result, she can refine the prompts � as Kashtanova did hundreds of times � until the image matches her intellectual conception. The more one uses prompts, the better one becomes at predicting the results. That�s why new jobs as prompt engineers have sprouted. Businesses seek the fruits of these creators� �intellectual labor.�"

The derivative nature of this creativity reminds me of the furor over Shepard Fairey's Hope poster (although that case was complicated by the fact that he lied about the source image in a legal action)

The article includes the by-now-common analogy to photography and discusses an 1884 supreme court case about a photograph of Oscar Wilde where the court concluded that photographs were not simple mechanical reproductions, but �representatives of original intellectual conceptions of the author.� I think the fallacy is in comparing the system behind generative AI to the equipment that takes a photograph (and even more off base if you include post-processing software that the creator of a photograph almost always uses. If generative AI systems were based on robots having gone out to paint and photograph Yosemite, the deep dark woods, etc., there'd be some fair comparison, but that's not what has been done.

The top three most liked comments on the article:

"Sorry dude. Copyright Office is right. You're wrong. Besides your analogies fail. A jazz artist is wholly creating their improvisations. A prompt "artist" is tweaking someone else's. In music that is called arranging not creating. Arrangements dont get copyrights."

"Exactly. I'm an engineer who does computer coding. If I ask chatGPT to create a code to, say, run a basic spectral analysis on gappy satellite altimetry data, using a specific non-FFT method, I absolutely SHOULD be allowed to USE the results (assuming they're right, which they might well not be for technical reasons, but I digress). But I surely shouldn't be able to COPYRIGHT that code. Because I didn't write it. It's literally made from a computer recombination of the work of other previous coders and mathematicians. It's not "mine", even if I can use it for free.

There's nothing sacred about art that makes it different than anything else. If you want to copyright it, make it yourself. Otherwise, be grateful that you can use the skill of other people (AI programmers and real artists) to get your work done well for you."

"And this is generally true of the current machine learning trained on big data model of AI. Every sentence, every image, comes from recombining the words and images the AI is trained on, a huge portion of which was nominally protected by copyright. Current AI models are the worst enemy of copyright in history."