“I never, never, never saw so many people in my whole life.”
This is what Jenny Lewis said after singing “I Never” about halfway through Rilo Kiley’s set Saturday night, and yeah, it makes sense. The last time the L.A. natives played their hometown was the summer of 2008, for what turned out to be their then-final show. The band split shortly after due to the romantic and creative tensions between Lewis and lead guitarist Blake Sennett. And with Lewis since moving onto a successful solo career, it felt likely to remain their final show.
So this reunion was a surprise to everyone, including Lewis, who at one point told the audience: “It’s amazing to be here with you all, but mostly…” She then gestured to her bandmates and said: “It’s amazing to be here with you all.”
They picked the right time, though. Their music remains hugely influential on both modern indie (Phoebe Bridgers) and pop (Olivia Roderigo) music, while fans from their original run are now in their 30s and 40s and ready to revisit the songs that provided a caustically witty soundtrack to the flirtations, relationships, and heartbreaks of their 20s.
To put it simply, Rilo Kiley was the clear draw of this year’s Just Like Heaven, the annual Pasadena music festival dedicated to the indie rock and electropop of the 2000s/early 2010s. Or, in terms of mood and aesthetics, making the Rose Bowl’s Brookside Park feel like the 2005 Coney Island Siren Music Festival. As someone in his mid-40s who loves indie rock (and was at those old Siren Festivals), every detail of the fest felt catered to me, from the food to the merch to a dance floor where the very same DJ I danced to every weekend in my 20s was spinning.
Bands started playing at noon, but the nearly 100-degree heat kept the daytime crowds thinner than years past. Which was a shame, because the temperature didn’t stop most artists from giving it all on stage. London’s Block Party played their first US show in years, with lead singer Kele Okereke showing (across hits like “Helicopter” and “This Modern Love”) that his voice can still go from sharp post-punk speak-singing to soaring theatricality on a dime.
Perfume Genius, a newer artist by this festival’s standards, swayed and sashayed with emotional vocals and and core-defying moves even during sun’s hottest hour. The also recently reunited TV On The Radio delivered a tight, propulsive set and (surprisingly) the only political messages of the day, with an intro of “Fuck these fascists comma fuck these fascists comma fuck these fascists comma fuck these fascists ” for “Trouble” and guitarist Kyp Malone’s “Somebody I Love is Palestinian” T-shirt.
(Credit: Ashley Osborn)
However, by 8:40pm, the sun was down and the crowds seemed to have doubled in size for Rilo Kiley. More than that, from the first notes of “The Execution of All Things,” the audience was re-energized and ready to dance. The band returned this energy in kind, as across the next hour, you’d never guess this group hadn’t played regularly together for years. From a brightly shimmering “Wires and Waves” to a darkly grooving “The Moneymaker,” they played songs from all their albums, even going back to their early cult hit “Frug,” complete with Lewis doing the song’s signature dance moves and the audience joining her bandmates for their backing vocals.
Numbers like this and the indie rock torch-song “I Never” were dramatically thrilling reminders of why Jenny Lewis was and still is the queen of her scene (hence her actual tiara). Her every gesture and expression made the audience go wild. She elicited cheers with a mere eyebrow raise on “Paint’s Peeling,” made hearts swoon with a kiss and middle finger during “Does He Love You?”’s explosive outro, and got the crowd singing for “With Arms Outstretched” to only then immediately silence them for the start of “A Better Son/Daughter.” Maybe it’s her past as an actor, but Lewis knows exactly what choice to make in each moment to work an audience.
This wasn’t a Jenny Lewis solo show, though, as her bandmates shined just as brightly. Sennet effortlessly commanded the crowd on each guitar solo and his lead vocal number “Dreamworld.” The lively interplay between him, bassist Pierre de Reeder, and drummer Jason Boesel was infectious, too. From the easy flow to the genuine smiles, all four members seemed to be having what you want most from a favorite band on their reunion tour: fun.
By the time they closed with their “our relationship is fucked, isn’t it?” anthem “Portions For Foxes,” Rilo Kiley’s set had been so electric and fulfilling that, honestly, it’s hard to figure out why they weren’t the headliners. Nothing was going to top them, especially not on their turf.
To see our running list of the top 100 greatest rock stars of all time, click here.
** Willie Nelson, Neil Young, John Mellencamp and Dave Matthews and Tim Reynolds will assemble Sept. 20 for the 40th anniversary of Farm Aid, which will take place Huntington Bank Stadium in Minneapolis. Billy Strings, Nathaniel Rateliff & the Night Sweats, Trampled by Turtles, Waxahatchee, Black Pumas‘ Eric Burton, Jesse Welles, Madeline Edwards and Margo Price will also perform.
“Family farmers are the heart of this country, and we depend on each other for good food and strong communities,” says Nelson, who helped found the organization in 1985. “For 40 years, Farm Aid and our partners have stood with farmers, supporting them to stay on their land even when corporate power, bad policies and broken promises make it harder to keep going. This year, we’re proud to bring Farm Aid to Minnesota to celebrate the farmers who sustain us and to fight for a food system that works for all of us. Family farmers aren’t backing down, and neither are we.”
** Mumford & Sons are reprising their 2011 Railroad Revival tour this summer for four outdoor shows in New Orleans (Aug. 3), Spartanburg, S.C. (Aug. 4); Richmond, Va. (Aug. 5) and Burlington, Vt. (Aug. 7). The venues for each are located near local train stations. Mirroring the concept of the original run, Mumford will be complemented by different guests at each gig, including Nathaniel Rateliff, Chris Thile, Trombone Shorty, Madison Cunningham and Lucius.
“We felt we had some unfinished business on the great American railroad,” says Marcus Mumford. “So, we’ve spent a lot of time cooking up this idea for another rolling festival to rip through the south and east of the U.S. The spirit of what we do, at its core, is always about people and collaboration. So, every show will be a collaborative performance from a bunch of our favorite people on the planet, and every show will be different. This will, without doubt, be the coolest house band we’ll ever get to play in.”
Tickets go on sale Friday (May 16). Mumford will also tour behind their newest album, Rushmere, in standard venues beginning June 5 in Bend, Or.
** Flying Lotus and Nubya Garcia have been added to the lineup for the Newport Jazz Festival, which will be held Aug. 1-3 at Fort Adams State Park in Newport, R.I. The event takes place the weekend after the Newport Folk Festival on the same grounds. Previously announced performers include Janelle Monae,the Roots, Jacob Collier, RAYE, Jorja Smith, Esperanza Spalding, the Yussef Dayes Experience. Preservation Hall Jazz Band, Kokoroko and De La Soul.
For Flying Lotus, Newport is the only upcoming show on his calendar beyond Friday’s (May 16) Cosmic Music: The Celestial Songs of Alice Coltrane at New York’s Carnegie Hall. There, Lotus, who is Coltrane’s grandnephew, will join Ravi Coltrane and several other family members to perform world premiere orchestral arrangements of her compositions.
** A stellar lineup of electronic music performers and DJs will head to Philadelphia’s Fort Mifflin from Sept. 19-21 for the long-running, DIY Making Time ∞ festival, now in its 25th year. Four Tet, ¥ØU$UK€ ¥UK1MAT$U, Boy Harsher, Panda Bear, and Moodymann will headline, with Ben UFO, Optimo (Espacio), Moor Mother, Suzanne Ciani, Windy & Carl, Laraaji, John Talbot,Avalon Amerson and VTSS also on board. Loidis and DJ Python will team for a B2B set, which will follow the former staging an ambient live show under his Huerco S alter-ego. Click here for all the details.
To see our running list of the top 100 greatest rock stars of all time, click here.
As AI continues to challenge the boundaries of creativity and copyright in the music industry, SoundCloud has found itself in the crosshairs of a growing conversation about trust, transparency and technology.
This past week, a clause buried in SoundCloud’s updated terms of service captured the artist community’s attention. First flagged by Futurism, the clause suggests that music uploaded to the platform could, in some cases, be used to “inform, train, develop or serve as input to artificial intelligence.”
The revelation is surfacing deep-seated anxieties among independent artists about how their content is being used behind the scenes. But SoundCloud, long considered a cornerstone of artist empowerment and grassroots discovery, is seeking to clarify intentions.
The company responded through a statement and unequivocally denied that its service has trained AI on the content of its users.
“SoundCloud has never used artist content to train AI models, nor do we develop AI tools or allow third parties to scrape or use SoundCloud content from our platform for AI training purposes,” a SoundCloud spokesperson told Futurism. “In fact, we implemented technical safeguards, including a ‘no AI’ tag on our site to explicitly prohibit unauthorized use.”
The company emphasized that its engagement with AI has been focused on enhancing user experience through tools such as personalized recommendations and fraud detection—not harvesting creative works to feed generative algorithms.
Still, the terms, updated in February 2024, arrived just as legal and ethical scrutiny around generative AI in music hit a fever pitch. Labels have been battling with tech firms over training data and artists have voiced concerns about their identities being synthesized without permission.
For now, SoundCloud’s message is one of reassurance, but the industry will be watching closely to see how that message holds up in practice. You can read their full statement below.
SoundCloud has always been and will remain artist-first. Our focus is on empowering artists with control, clarity, and meaningful opportunities to grow. We believe AI, when developed responsibly, can expand creative potential—especially when guided by principles of consent, attribution, and fair compensation.
SoundCloud has never used artist content to train AI models, nor do we develop AI tools or allow third parties to scrape or use SoundCloud content from our platform for AI training purposes. In fact, we implemented technical safeguards, including a “no AI” tag on our site to explicitly prohibit unauthorized use.
The February 2024 update to our Terms of Service was intended to clarify how content may interact with AI technologies within SoundCloud’s own platform. Use cases include personalized recommendations, content organization, fraud detection, and improvements to content identification with the help of AI Technologies.
Any future application of AI at SoundCloud will be designed to support human artists, enhancing the tools, capabilities, reach and opportunities available to them on our platform. Examples include improving music recommendations, generating playlists, organizing content, and detecting fraudulent activity. These efforts are aligned with existing licensing agreements and ethical standards. Tools like Musiio are strictly used to power artist discovery and content organization, not to train generative AI models.
We understand the concerns raised and remain committed to open dialogue. Artists will continue to have control over their work, and we’ll keep our community informed every step of the way as we explore innovation and apply AI technologies responsibly, especially as legal and commercial frameworks continue to evolve.
Sonic Temple 2025 is in the books, but you can relive the good times and enjoy the experience through four photo galleries from the music weekend. Continue reading…
Planet 13, the world’s largest cannabis dispensary, is poised for what its executives are calling the busiest weekend of the year: not New Year’s Eve and not 4/20, but EDC.
The cannabis giant is celebrating the festival’s return to Las Vegas by giving away 20 three-day passes to lucky visitors as it launches a slate of EDC-themed offerings in collaboration with RNBW, the Insomniac Events-affiliated cannabis brand.
“EDC has become the busiest weekend of the year at Planet 13 Las Vegas,” said Bryant Ison, Planet 13’s Vice President of Marketing. “It’s our Black Friday, busier than 4/20, Super Bowl and New Year’s Eve. We see a 40% increase in revenues and close to a 30% increase in customer traffic at our flagship Vegas SuperStore during EDC.”
The promotion, part of the dispensary’s “Daisy Daze” weekend, features a full-scale retail activation with live vendor pop-ups, food trucks, limited-edition merch and specially curated cannabis bundles. Hopeful festivalgoers receive raffle entries for every bundle purchased and winners will walk away with passes to EDC, North America’s largest EDM festival.
The connection between Planet 13 and EDC Las Vegas runs deeper than seasonal marketing, however. RNBW is a luxury cannabis label co-developed through a strategic partnership between EDC’s organizer, Insomniac, and industry stakeholders. RNBW’s “Dark Rainbow” strain, available exclusively at Planet 13, is the dispensary’s featured drop this year.
Pasquale Rotella, Insomniac founder and CEO, has championed RNBW as part of a broader initiative “intersecting live music and cannabis culture,” he said in a 2021 statement.
EDC returns to the Las Vegas Motor Speedway this weekend from May 16-18 with Martin Garrix, Alesso, Tiësto, REZZ and many more. You can find the festival’s full lineup and read more about its 2025 edition here.
Republican lawmakers slipped language into the Budget Reconciliation Bill this week that would ban AI regulation, on the federal and state levels, for a decade, as 404 Media reports.
An updated version of the bill introduced last night by Congressman Brett Guthrie (R-KY), who chairs the House Committee on Energy and Commerce, includes a new and sweeping clause about AI advancement declaring that “no State or political subdivision thereof may enforce any law or regulation regulating artificial intelligence models, artificial intelligence systems, or automated decision systems during the ten year period beginning on the date of the enactment of this Act.”
It’s a remarkably expansive provision that, as 404 notes, likely reflects the engraining of Silicon Valley figures and influences into Washington and the White House. Tech CEOs have vied for president Donald Trump’s attention since he was inaugurated, and the American tech industry writ large has become a fierce and powerful lobbying force. The Trump administration is also stacked with AI-invested tech moguls like David Sacks, Marc Andreessen, and Elon Musk.
Meanwhile, the impacts of a regulation-free AI landscape are already being felt. Emotive, addictive AI companions have been rolled out explicitly to teenagers without evidence of safety, AI companies are missing their climate targets and spewing unchecked emissions into American neighborhoods, and nonconsensual deepfakes of women and girls are flooding social media.
No regulation will likely mean a lot more fresh hell where that came from — and little chance of stemming the tide.
New reconciliation bill text in the House would ban states from enacting any laws on AI for a decade.
As state lawmakers work to address impacts on minors, on workers, on data privacy, and more, this could have devastating consequences. pic.twitter.com/7Ubbe4ltQX
— Americans for Responsible Innovation (@americans4ri) May 12, 2025
Blank Checks
The update in the proposed law also seeks to appropriate a staggering $500 million over ten years to fund efforts to infuse the federal government’s IT systems with “commercial” AI tech and unnamed “automation technologies.”
In other words, not only does the government want to completely stifle efforts to regulate a fast-developing technology, it also wants to integrate those unregulated technologies into the beating digital heart of the federal government.
The bill also comes after states including New York and California have worked to pass some limited AI regulations, as 404 notes. Were the bill to be signed into law, it would seemingly render those laws — which, for instance, ensure that employers review AI hiring tools for bias — unenforceable.
As it stands, the bill is in limbo. The proposal is massive, and includes drastic spending cuts to services like Medicaid and climate funds, slashes that Democrats largely oppose; Republican budget hawks, meanwhile, have raised concerns over the bill’s hefty price tag.
Whether it survives in its current form — its controversial AI provisions included — remains to be seen.
On Friday, the US Copyright Office released a draft of a report finding that AI companies broke the law while training AI. The next day, the agency’s head, Shira Perlmutter, was fired — and the alarm bells are blaring.
The report’s findings were pretty straightforward. Basically, the report explained that using large language models (LLMs) trained on copyrighted data for tasks like “research and analysis” is probably fine, as “the outputs are unlikely to substitute for expressive works used in training.” But that changes when copyrighted materials (like books, for example) are used for commercial applications — particularly when those applications compete in the same market as the original works funneled into models for training. Other examples: Using an AI that gets trained on copyrighted journalism, in order to create a news generation tool, or using copyrighted artworks, in order to then create art to sell. That type of use likely breaches fair use protections, according to the report, and “goes beyond established fair use boundaries.”
The report’s findings seem to strike a clear blow to frontier AI companies, who have generally taken the stance that everything ever published by anyone else should also be theirs.
OpenAI is fighting multiple copyright lawsuits, including a high-profile case brought by The New York Times, and has lobbied the Trump Administration to redefine copyright law to benefit AI companies; Meta CEO Mark Zuckerberg has taken the stance that others’ content isn’t really worth enough for his company to have to bother compensating people for it; Twitter founder Jack Dorsey and Twitter-buyer-and-rebrander Elon Musk agreed recently that we should “delete all IP law.” Musk is heavily invested in his own AI company, xAI.
Clearly, an official report saying otherwise, emerging from the US federal copyright-enforcement agency, stands at odds with these companies and the interests of their leaders. And without a clear explanation for Perlmutter’s firing in the interim, it’s hard to imagine that issues around AI and copyright — a clear thorn in the side of much of Silicon Valley and, to that end, many of Washington’s top funders — didn’t play a role.
As The Register noted, after the report was published, legal experts were quick to catch how odd it was for the Copyright Office to release it as a pre-print draft.
“A straight-ticket loss for the AI companies,” Blake. E Reid, a tech law professor at the University of Colorado Boulder, said in a Bluesky post of the report’s findings.
“Also, the ‘Pre-Publication’ status is very strange and conspicuously timed relative to the firing of the Librarian of Congress,” Reid added, referencing the sudden removal last week of now-former Librarian of Congress Carla Hayden, who was fired on loose allegations related to the Trump Administration’s nonsensical war on “DEI” policies.
“I continue to wonder (speculatively!),” Reid continued, “if a purge at the Copyright Office is incoming and they felt the need to rush this out.” Reid’s prediction was made before the removal of Perlmutter, who was named to her position in 2020.
To make matters even more bizarre, Wired reported that two men claiming to be officials from Musk’s DOGE squad were blocked on Monday while attempting to enter the Copyright Office’s building in DC. A source “identified the men as Brian Nieves, who claimed he was the new deputy librarian, and Paul Perkins, who said he was the new acting director of the Copyright Office, as well as acting Registrar,” according to the report.
The White House has yet to speak on why Perlmutter was fired, and whether her firing had anything to do with Musk and DOGE. It wouldn’t be the first time, though, that recent changes within the government have benefited Musk and his companies.
Dan Schwarz, CEO of Futuresearch, recently shared insights from his company’s ongoing analysis of OpenAI and the broader Generative AI market. Futuresearch has recently focused on dissecting OpenAI’s revenue composition to forecast its growth prospects, publishing several analytic reports on the topic. What follows is a heavily edited excerpt from that conversation, covering both recent findings and previously unpublished projections from Futuresearch’s research.
What is your headline takeaway from your analysis of OpenAI’s revenue projections?
Futuresearch was the first to reverse-engineer OpenAI’s revenue streams before they were publicly disclosed, and we’ve been tracking them for over a year. OpenAI’s projection of $125 billion by 2029 is plausible in theory but highly implausible in practice. This relates to a recent report called AI 2027 that describes a scenario where a frontier lab experiences a runaway AI takeoff based on certain revenue projections. When we calibrate OpenAI’s projections against our expert forecasts, we find that hitting these numbers would require unprecedented exponential growth that doesn’t align with observed data or competitive realities.
What is your alternative revenue projection for OpenAI, and why is the range so wide?
Our 90% confidence interval for OpenAI’s 2027 revenue spans roughly $10 billion to $90 billion. This is an extraordinarily wide range because OpenAI is perhaps the most uncertain business possible to forecast. On one hand, they could monopolize multiple industries through rapid exponential growth. On the other hand, they could stumble due to ongoing litigation, talent exodus, and competition from other labs that already have better models in some categories. I personally lean toward the more bearish end of our internal forecasts. The uncertainty grows substantially when extending forecasts beyond 2027, making the 2029 projection even more speculative.
[Note: This forecast was later updated, reflecting new considerations; the revised figures ($11B −$70B, median $41B) are presented in the graphic below.]
How is OpenAI’s revenue currently split between different sources?
Contrary to early assumptions, API calls account for no more than 15 percent of OpenAI’s revenue as of mid-2024. The bulk comes from ChatGPT’s consumer tier and increasingly from ChatGPT Enterprise. When we published our first analysis, many people thought API was the dominant source, but we demonstrated that ChatGPT was driving most of their revenue, and this pattern continues today.
Ready for more? Upgrade to premium for exclusive content!
What data sources and methods underpin your forecasts?
A good forecast blends data extrapolation with judgmental forecasting, adjusting for factors not captured in historical data. We start by extrapolating OpenAI’s initial revenue ramp (from $1B to $3.5B annually), but this provides only a crude baseline given the limited data points.
More importantly, we track competition closely by evaluating how good OpenAI’s models are compared to alternatives (Claude, Gemini, Llama, DeepSeek, etc.). This is challenging because benchmarks change constantly and don’t necessarily reflect actual user experience or specific use cases.
We also use judgmental forecasting techniques similar to prediction markets or Tetlock’s work for questions with limited direct data, such as lawsuit outcomes. This approach naturally yields wide intervals rather than spurious precision.
How do OpenAI’s growth projections compare to other tech giants historically?
Looking at historical data: Microsoft took 28 years to reach $100 billion in revenue, Amazon took 18 years, Google took 14, Facebook took 11, and ByteDance just 6. When you model a roughly 40% decade-over-decade acceleration, it implies that a frontier lab could theoretically reach $100 billion within four years—consistent with the AI 2027 scenario—but that would require outpacing even ByteDance by a wide margin. Achieving that level of growth would require monopolistic dominance, which is far from guaranteed given the competitive landscape.
Why are you skeptical that ChatGPT can continue as the primary revenue driver?
I don’t believe in ChatGPT as a long-term driver for massive revenue. It faces the most competitive pressure. Free tiers from Google (Gemini), Meta (Meta.ai), Anthropic, and others are already excellent, sometimes better for specific use cases, and often multimodal. Meta, in particular, is aiming squarely at ChatGPT, potentially making it an open-source commodity.
It’s hard to imagine tens of millions of people paying $20/month long-term when comparable or better free alternatives exist. While subscriber numbers were estimated around 23 million paying users (as of April 2024), churn is reportedly high. Every time you use ChatGPT, they’re likely losing money due to inference costs. Both Google and Meta have massive war chests compared to OpenAI, which has had to raise stupendous amounts of venture capital just to get this far.
What about API revenue? Can that become a significant growth driver?
I don’t believe in the API as a source of massive, exponential growth either. The API market is intensely competitive – it’s a race to the bottom on price. Google appears to be winning on the frontier of quality-for-cost right now. Models from Google, Anthropic, and open-weight options (Llama, DeepSeek) are excellent and often cheaper or better for specific needs.
Customers can switch providers “on a dime” – it’s easy to stop spending with OpenAI and move to Google or Anthropic, as my own company has done. The idea of API revenue growing into a “ten billion behemoth” seems very implausible. This dynamic makes it hard for OpenAI to preserve margins on either API or consumer tiers.
If neither ChatGPT nor API will drive massive growth, what could?
Agents represent the only credible path to scaled revenue that I find plausible, though still uncertain. If OpenAI reaches the higher end of our revenue projection ($60-90 billion by 2027), at least one-third would likely come from agent-based revenue – and they barely generate any agent revenue right now.
Success would hinge on OpenAI launching products specifically for software automation and ramping them to billions in revenue within 1-3 years. This could involve automating complex white-collar work through systems like their “operator” concept that can control a computer to perform tasks. Examples include financial analysis, software automation, or general computer operation.
The recent release of o4, despite delays, represents a step-function improvement in agentic flows, particularly web research. It shot to the top of our leaderboards for complex tasks requiring reasoning, tool use, and overcoming gullibility, surpassing even Gemini 2.5 Pro and Claude models on those specific tasks at that time.
Does OpenAI have a sustained technical advantage over competitors?
No – there’s no single “head and shoulders” leader. The advantages appear extremely fleeting right now. An edge gained one month (like GPT-4o in web research) could be lost the next. DeepSeek in China, Meta, Google – everyone is iterating rapidly.
The definition of “foundational model” gets tricky. OpenAI would need a decisive advantage in the capability that drives revenue. If agents are the key, they need the best agentic capabilities. o4 looks like reinforcement learning applied over a base model to perform tasks (tool use, search, code execution). Is the underlying base model the best, or is the agentic layer on top the key differentiator? It’s not entirely clear.
Unless a lab achieves a true, defensible breakthrough, the current state feels more like a continuous leapfrogging race where leadership changes frequently. This makes long-term revenue projections based on current leads very fragile.
What’s the path to a potential “winner takes all” dynamic in AI?
The most plausible path to that kind of scenario involves a positive feedback loop in research automation. If any frontier lab (not necessarily OpenAI – could be DeepMind, Meta, Anthropic, X.ai) can significantly automate its own software engineering and research processes (coding, running experiments, analyzing results), its researchers become vastly more productive.
If they can make research 3x or 10x faster, they could gain an insurmountable advantage. They use that advantage to further automate their research, getting faster and faster. This positive feedback loop is where a winner-takes-all dynamic could emerge, leading one company to pull years ahead of competitors who were previously only months behind. This seems to be the path towards the kind of monopolistic advantage OpenAI would need, and labs are likely working on this explicitly.
How significant is the talent exodus from OpenAI?
The talent exodus from OpenAI is an underrated problem. The number of great researchers who have left to directly compete with OpenAI is probably unprecedented for a leading tech company. We’ve tracked key departures – most have gone to competitors like X.AI, Anthropic, and Ilya Sutskever’s Safe Superintelligence.
In AI, having the right brilliant people in the right place might be the deciding factor. Companies like Anthropic are attracting top talent from both Google and OpenAI, and I don’t see movement in the opposite direction. When brilliant graduates from top CS PhD programs choose between offers from frontier labs, there’s a good chance they might choose Anthropic over OpenAI, which could prove to be a decisive advantage in the long run.
How important are multimodality, coding, and robotics for future revenue?
Multimodality: For web-research agents, multimodal abilities (reading screenshots, extracting tables, parsing infographics) are crucial. However, Google appears to be leading in this area currently. Multimodal capabilities are critical for agents working with knowledge workers, including customer service agents who need to handle calls, talk, listen, and potentially join video calls.
Coding: This represents a multi-trillion-dollar opportunity that doesn’t strictly require multimodality. If OpenAI could create armies of superhuman coders, that’s a path to revenue. However, they aren’t clearly the best now – engineers have flocked to Claude for coding, while Gemini and GPT models remain competitive.
Robotics: Despite decades of high expectations, warehouse and manufacturing automation remain largely manual. Amazon’s robotics is limited, and Boston Dynamics has yet to deliver widely adopted commercial robots. Modern generative-AI techniques might ignite a new robotics wave, but history suggests we should brace for potential disappointment over the next decade. The analogy to self-driving cars is relevant – Tesla’s Full Self-Driving promises haven’t materialized as advertised, while Waymo’s more incremental approach led to slower but real deployment.
futuresearch’s May 2025 Deep Research Bench: ChatGPT-o3+search (default o3, not OpenAI Deep Research) leads in agentic web research. Click HERE to enlarge.
How does Anthropic compare in this landscape?
Anthropic faces similar challenges to OpenAI but with a more concentrated risk profile, being heavily dependent on API revenue, much of which comes via AWS. As API becomes increasingly competitive with pricing pressure, this creates significant vulnerability.
However, Anthropic has potential advantages. They focus heavily on interpretability – understanding and tweaking the internals of their models – possibly more than any other lab. If they can make breakthroughs in understanding why these models are so capable and how to better align them, they could gain a decisive technical edge.
Anthropic also seems to be winning in talent acquisition. Many brilliant researchers who left Google and OpenAI have gone to Anthropic. If safety and reliability become critical differentiators – if other models start engaging in problematic behaviors – Anthropic’s focus on alignment could become a major competitive advantage.
What’s the key takeaway for teams building AI applications?
Don’t assume OpenAI’s current or projected dominance is guaranteed. The market is fiercely competitive, and advantages are temporary. Design your systems to be model-agnostic; avoid locking yourself into a single provider.
Experiment with models from Google, Anthropic, Meta (Llama), DeepSeek, and others – you might find better performance or cost-effectiveness for your specific use case. Be prepared for rapid shifts in capabilities and pricing.
While OpenAI could achieve massive success through agents or a research breakthrough, their path is far more uncertain than their projections suggest. Focus on the practical utility and cost of different models for your application today, while keeping an eye on the potential for disruptive agentic capabilities in the near future from any of the major labs.
