Good morning. Spotify (and Apple Music) wrapped came out this week (at around the same time as a report, which we get into below, that music creators are expected to lose a quarter of their revenue due to GenAI by 2028).

I clocked 58,784 minutes across nearly 1,000 artists this year. Tons of new discoveries, but Judah & The Lion, Marianas Trench and The Lone Bellow topped my charts.

(Sorry we’re a little late today, technical difficulties)…

— Ian Krietzberg, Editor-in-Chief, The Deep View

In today’s newsletter:

• 🔭 AI for Good: Space weather

• 📊 November was the biggest month on record for AI investment

• 💰 Forecast: Creators to lose a quarter of revenue to GenAI by 2028  

• ⚡️ Meta and Big Tech’s nuclear revolution

AI for Good: Space weather

The New Jersey Institute of Technology (NJIT) in October was awarded a $5 million NASA grant to fund a new research center focused on developing AI-powered solar eruption prediction tools. 

The details:  The new center will partner with NASA, New York University and IBM to advance artificial intelligence and machine learning tools to predict different kinds of solar eruptions, including solar flares and coronal mass ejections (CMEs). 

• These eruptions — a few of which happened this year — can create geomagnetic storms that, when they hit Earth, can disrupt satellites and earth-bound power grids alike. 

• But, according to NJIT, a lack of data around the mechanisms that lead to these eruptions has historically made it difficult for astronomers to develop accurate predictions; the new center plans to integrate advanced machine learning and AI technology with NASA’s solar observation posts in an attempt to increase our understanding of these murky mechanisms. 

Why it matters: Haimin Wang, a distinguished physics professor at NJIT who will lead the project, said: “By harnessing AI-enabled tools to investigate the fundamental nature of space weather, we aim to significantly enhance our ability to interpret observational data from the Sun to forecast major solar eruptions accurately and in near real-time, a capability beyond our reach up to this point.”

Finally, a developer tool that makes digging around for the answer a thing of the past.

Other developer tools can’t tell you how your codebase works and why. Unblocked can. We augment your source code with context from Slack, Confluence, Jira (and more), so your team gets helpful answers instantly.

Unblocked gives you:

• Instant answers about every aspect of your codebase 

• Relevant, historical context for any code open in an IDE

• Automated answers to questions asked in Slack, removing interruptions to you

⌛Engineers who use Unblocked save an hour or more a day.

“Unblocked empowers every developer with the insight once reserved for the seasoned few. ” Mathew Spolin - SVP Engineering at AppDirect

Start your 21-day trial today

November was the biggest month on record for AI investment

Venture funding around the world hit $28 billion last month, according to Crunchbase data, a decline over the $32 billion notched in October, but a significant increase compared ot the $21 billion recorded last year. 

And more than half of that number went to AI companies. 

The details: $14.5 billion in venture capital dollars went to companies that sit beneath the broad AI umbrella, including robotics, autonomous driving, security, healthcare and marketing, in addition to the expected developers and chipmakers. 

• This number, the largest monthly number on record for AI investment, represents an increase over the $12.2 billion invested in October, the $4.3 billion invested in September and the $5.5 billion invested in August. 

• But, unlike previous record numbers, this number is stretched across the thinnest number of deals on record: a mere 265 (October saw 406 and September saw 513, according to Crunchbase data). 

The big raises last month came from xAI and Anthropic — xAI raised some $5 billion at a $50 billion valuation; Anthropic raked in $4 billion (all from Amazon). 

It comes at a time, according to Bain Capital Ventures partner and AI investor Rak Garg, when VCs have to “be smart about the companies you invest in.” But despite the somewhat tumultuous nature of the AI-related public-market investments, VC excitement — evidenced in part by these November numbers — has continued unabated for months. 

This is your last chance to invest in the startup that's transforming AI

The AI revolution is well underway, with generative AI becoming an ever more popular product. But AI isn't magic; it relies on data centers to run like the cloud.

And as the tech gets more popular, these data centers are simply running out of storage — more than 90 zettabytes of data will need to be created and stored in 2025 to keep pace with GenAI.

That's why Atombeam’s tech — which can reduce data size up to 75% — has investors jumping at the chance to join in. 

Atombeam's patented (AI-powered) Neurpac software can make networks up to 4x faster and more secure — helping clients avoid billions in expensive hardware upgrades.

Over 7,500 investors have joined Atombeam and their last round had an over $3M waitlist to invest. Don’t miss your chance to get in on the startup changing the paradigm on how we store and send data - faster, encrypted, and more secure. 

Plenty of industry players have taken note of Atombeam's approach; in addition to partnering with NVIDIA, Ericsson, and Intel., the company has $2.4M contracts in place with the U.S. Air Force and Space Force to develop Neurpac for use in military satellites.

Atombeam’s raise just hit $10M - but this round is ending in only 13 days! Become an Atombeam shareholder for just $8/share before this round closes.

• Misinformation researcher admits ChatGPT added fake details to his court filing (The Verge).

• Apple hits snags adapting Baidu’s AI models for China users (The Information).

• OpenAI partners with defense company Anduril (CNBC).

• South Korean president faces impeachment after shock martial law move (Semafor).

• Companies in Mexico embrace AI to resurrect the dead (Rest of World).

If you want to get in front of an audience of 200,000+ developers, business leaders and tech enthusiasts, get in touch with us here.

• Appearing at the New York Times’ annual DealBook Summit, OpenAI CEO Sam Altman said a few things of interest, notably that the “12 Days of OpenAI” — a period of daily livestreams featuring new demos — will begin tomorrow. He also said that ChatGPT has reached 300 million weekly active users.

• Amazon has released a new set of foundation models — nicknamed Nova — that can process text, images and video. All models will be available on Amazon Bedrock.

Forecast: Creators to lose a quarter of revenue to GenAI by 2028  

Musicians around the world are expected to lose 24% of their revenue to generative AI by 2028, according to a global economic study that sought to examine the impact of generative AI on creative sectors. 

Additionally, those who work in the audiovisual industry are expected to lose some 21% of their revenue to GenAI by 2028, losses that come as the market for AI-generated music and audiovisual content is expected to surge from $3.1 billion today to $67 billion four years from now. 

The details: The study was conducted by the International Confederation of Societies of Authors and Composers (CISAC), which represents 5 million creators around the world. 

• By 2028, according to the report, the annual market size of GenAI in music is expected to hit $16.8 billion; that same year, creators are expected to lose $4.2 billion. 

• By 2028, the report predicts that GenAI music “will account for around 20% of music streaming platforms’ revenues and around 60% of music libraries revenues.”

Over the next five years, just in these two categories, a cumulative total of some $23.1 billion in revenue for creators is at stake, according to the report: “In an unchanged regulatory framework, creators will actually suffer losses on two fronts: the loss of revenues due to the unauthorized use of their works by Gen AI models without remuneration; and replacement of their traditional revenue streams due to the substitution effect of AI-generated outputs, competing against human-made works.”

CISAC President Björn Ulvaeus said in a statement that the study is an important guideline for policymakers around the world; and indeed, some are definitely thinking about this. Just last week, an Australian senate committee released a report that called out the “unprecedented theft” of artists’ work, and called for a strong regulatory framework to address it. 

Meta and Big Tech’s nuclear revolution

Meta is working to identify nuclear power developers to help the company meet its “AI innovation and sustainability objectives.” The social media giant is looking to add 1-4 gigawatts of new nuclear energy capacity in the U.S. (based on average numbers in the U.S., just one gigawatt is enough to power 876,000 households for an entire year). 

The details: Meta said in a post that advancements in AI innovation require the electrical grid to expand by embracing “reliable, clean and renewable energy,” something that nuclear offers. 

• Meta said it hopes these new nuclear plants will be operational by the early 2030s. 

• “As we look ahead to our next decade of innovation and growth, we are planning for our data center energy needs while simultaneously contributing to a reliable grid and advancing our sustainability commitments,” Meta said. 

The context: The Big Tech nuclear revolution is now fully in focus. In recent months, Microsoft, Google and Amazon all announced projects to bring nuclear power online over the next decade.

• There is a clear synchronicity in their reasoning here: AI is increasing their energy needs in such a dramatic fashion that the grid will be unable to support it; nuclear, they have decided, is the best option. 

• The push to nuclear comes as investments in AI have pushed all the Big Tech firms further from achieving their sustainability goals, due to the enormous energy cost, water consumption and carbon emissions of the data centers that power these energy-hungry products. 

The problem with nuclear: The idea behind nuclear is that it is stable, whereas other green energy options — like wind and solar — are intermittent.  

But, according to Mark Jacobson, a professor of civil and environmental engineering at Stanford, nuclear isn’t a good answer. First, there is a major lag time between inception and production (far longer than the two to five years for solar and wind projects); nuclear power is also several times more expensive than solar and wind and there remain risks around the radioactive waste produced by nuclear power plants. 

The lag time is so long that grids will continue releasing carbon emissions until a plant gets up and running, creating a massive opportunity cost that could have been addressed much more quickly with massive investments in solar and wind, whose stability issues can be addressed with battery storage techniques (in most of these Big Tech cases, we’re talking close to a decade). 

• Jacobson wrote that China’s investment in nuclear took so long to materialize that its grid emissions increased by 1.3% between 2016 and 2017 as a result, rather than declining by the expected average of 3%.  

• Current battery storage technology can store renewable energy for a minimum of 12 hours, depending on the type and size of the battery and the energy needs of the user. But non-battery storage technologies — long duration energy storage — exist as well; new research into LDES highlighted the importance of combining both technologies to unlock a zero-emission grid. 

Meta on Wednesday said it is building a $10 billion AI data center in Louisiana. It will be Meta’s largest data center in the world. 

The acknowledgment of the need for clean energy is good to see. But the push to nuclear — while not itself a terrible thing — feels misplaced and is surely moving far too slowly. For years, we have very desperately needed very low-emission energy solutions, not just to power additional energy demand, but to eat into and significantly reduce established fossil fuel use. 

That is not happening at the scale that we need it to. 

And the construction of nuclear plants to power the additional energy needs of AI means that existing fossil fuel-powered demand will remain untouched. We need to do more, and we need to do it faster. 

And at the cost of the infrastructure here … I would be far happier to see these massive companies invest billions in enhancing renewable battery storage technology, than revitalizing nuclear infrastructure (which, by the way, is not a zero-emission energy source). 

🤔 Your thought process:

Selected Image 2 (Left):

• “The rightmost hand is in an unnatural position to be holding the cup.”

💭 A poll before you go

Thanks for reading today’s edition of The Deep View!

We’ll see you in the next one.

Here’s your view on AI/cybersec training:

40% of you said their company has not required some form of anti-phishing or AI-related cybersecurity training; 36% said their company has.

Yes:

• “The education sector is a highly vulnerable workplace — the human element is immensely varied and diverse, reducing the effectiveness of "sheep-dip" training. It is necessary but insufficient to raise staff and faculty digital competence capabilities.”