We are SO back

📰 News Roundup

• OpenAI gets $500 Million for infrastructure: Sam Altman raises an eye-watering $500 Million - kind of, and billionaires bicker on X about whether the money exists. “I realize what is great for the country isn't always what's optimal for your companies” Altman fires at Musk… another cage match is brewing.

• Jesse Eisenberg is using AI to understand his anxiety when ordering bagels: 'Do you think if I had more circumstantially difficult life experiences, do you think I'd be more confident asking for a bagel?' he asks the world’s most expensive supercomputer.

• AI wins a Grammy: The Beatles win another Grammy with some help from AI. The previously unreleased track “Now and Then” uses machine learning to clean up the track, separating Lennon’s piano and vocals from a low-fi demo. The first ever song with AI assistance to win a Grammy.

👾 Interview: Moin Roberts-Islam

Meet Moin Roberts Islam, an AI and technology disruptor that has been an innovator in the industry for over a decade. Moin is a writer for Forbes, Innovation manager at the Fashion Innovation Agency, and is a highly respected voice on AI. We caught up with Moin on his AI career, and the rapidly evolving situation with DeepSeek.

As a Forbes reporter, covering AI and technology. What are you looking for from AI startups and technology, how do you separate meaningful advancements from features or investor fluff?

The innovations that excite me the most, are the ones that I would be excited to use or collaborate with myself.  Experimenting within tech, I have run up against recurrent problems, bottlenecks or grey areas many a time, so when a solution presents itself that goes towards answering, or at least mitigating, one of these areas, I get a genuine sense of excitement.

Beyond that, exciting new tools that will have an impact at a societal level, like a new wave of mixed-reality headsets, or a new suite of AI tools, are a good bet for a good story, since this presents something that a tech-leaning audience can easily relate to or imagine using.  Another metric that I use for assessing new technologies - “is it significantly better than, or different to, what is already out there?”  This still makes it relevant to a wider audience, since these kinds of technologies have the potential to displace incumbents. 

DeepSeek has dominated the news, with contrasting information fueling speculation. What is the likelihood that the $5.6 Million training figure is accurate and holds true to being able to train a model of that level for essentially peanuts?  

When I initially heard about this, I was open-minded to the possibility, and I was kind of rooting for the underdog.  We’re still in the early stages, relatively speaking, of developing new AI models and how to handle the associated computing and processing.  For the longest time, we’ve held the belief that the only way to improve performance is to throw more computational power behind something, but it’s encouraging to hear people deviating from this linear model of thinking. 

There is a huge amount of talent in China, no doubt - just look at the success of so many other tech businesses that have developed there.  A quick glance at the App Store will tell you everything you need to know - DeepSeek, Temu, TikTok, they’re all right up there. 

Having said that, recent reports have cast a lot of doubt on the veracity of the $5.6 million claims.  Compelling research from SemiAnalysis, amongst others, would suggest that the actual costs were many hundreds of millions higher than this, potentially running into billions, and that the figure of $5.6M only accounts for the cost of the final training run.  There has been conjecture that the hardware costs alone would have run as high as many hundreds of millions, with the R&D costs taking this close to an estimated $1.3 billion.

It was thought that OpenAI were creating a potential monopoly, but with the DeepSeek situation, do you think we’ll see a shift in AI towards a more decentralized AI ecosystem?

I firmly believe that we will have competing solutions being provided by AI companies from all across the world.  We have some of the largest and most highly educated populations in the world in both China and India, and it is absolute folly for us to think that they will not be able to produce solutions to compete with their counterparts in the West.  Likewise, there are many other countries around the world that have been investing in their education systems for years, specifically around computing and technology, so we will see lots of players and local markets bubbling up.

Add to this the availability of many open source AI models and pipelines - there is scope for a multitude of smaller, more specialised providers to thrive and add genuine value to their own customer bases.  Bigger is not always better.  We don’t all shop with Amazon or Shein, just because they have vast and impressively efficient logistics and supply chains.  So it also doesn’t make sense that we will go to OpenAI or any other behemoth for all of our AI needs.  

How does that impact the AI space from an innovation and investment standpoint?

Hopefully it blows the whole thing wide open.  For too long, we have seen certain demographics and certain geographies succeed more than others when it comes to investment, but hopefully this leveling of the playing field will allow for newcomers from all over to come in and pitch innovative new businesses to investors - they just need to be just as innovative in their approach to be able to see it in time.  Local providers for local markets, niche solutions, new ways of thinking and approaching problems - these should be the things that we are focusing on and looking at as markers of potential success, as opposed to the old adage of investing in the traditional “stale, pale and male” business owners.

Read more of Moin Roberts-Islam’s expertise on Forbes

📅 Upcoming Events

We have some big announcements coming soon. 🤫 Stay tuned NY - April 24th, LA - June 5

🔦 Spotlight

AE Studio
While everyone’s debating whether AI can properly order a bagel, AE Studio is quietly solving the real challenges. From helping vulnerable families navigate finances to powering Fortune 500 innovations at Berkshire Hathaway and Samsung, we’re focused on making AI actually deliver value. Think your company’s ready for what’s coming? Take 2 minutes to find out where you really stand in 2025's AI landscape - and what to do about it.

🧠 Learn More [HERE]

ArentFox Schiff Law
In an ArentFox Schiff report this month, the firm warns of legal complication from the new rapid trend of AI Agents. Putting a spotlight on developers of AI agents for future complications stating:

“Imagine asking ChatGPT or Gemini not just to develop a travel itinerary, but also to book your flights, rental car, and accommodations, if an AI agent inadvertently books the wrong flight, who is liable for the fare? How developers balance the strong desire for powerful automated agents against the safeguard of robust human supervision will be a defining choice for many.”

👉 Read the full report: [HERE]

🛠 Tool of the Week: Boardy

Boardy is Calling you!
Ever wanted a super-connector working for you night and day? Bored of endless networking events, and name tag wearing after-work socials? Just call Boardy… This great little guy learns who you are through a phone call, then gets to work connecting you to high level VCs, founders and more all through LinkedIn.

🤙 Call Boardy: [HERE]

Til next time,

Dani Van de Sande (Founder), James Joseph (The Weekly’s Editor) & the Artist and the Machine team.

ARTIST AND THE MACHINE is a free newsletter featuring the latest in AI x culture. Subscribe so you never miss an issue, and if you liked it, share with 1-2 friends .

xx,

Want to get in touch? Email us, follow/DM Dani on Twitter.