Scientists have identified what they believe is an accessible tunnel or cave on the moon. Such a feature, if real, may well define years of development by startups, governments and space companies aiming to create a lasting lunar colony.
Italian astronomers led by Leonardo Carrer and Lorenzo Bruzzone, working with Capella Space and JHUAPL, analyzed data collected by the Lunar Reconnaissance Orbiter in 2010. Specifically, its radar imagery of the surface, which in one place the team proposes was reflected in such a way that implies the presence of “a subsurface cave tens of meters long” — but potentially much larger.
It’s still at the bottom of a pit around 100 meters deep, but we’ll figure that part out later. What matters here is that we appear to have discovered a highly stable (otherwise it would have collapsed some time in the last few million years) moon cave that could serve as a base for lunar operations. Their findings are published in Nature Astronomy.
You may well ask: Why would we bother hiding in some ancient lava tunnel? Don’t we want to just make a surface habitat?
That’s certainly what we envision from science fiction, yes, but the reality is that the moon’s surface is quite an inhospitable place. With no ionosphere, it has no protection from solar or cosmic radiation, and its pocked surface shows how frequently it is bombarded by meteorites large and small, which are not slowed or burned up in an atmosphere. The temperature also varies from deep-space low to dangerously hot.
These and other factors mean that any surface dwelling would need to be extremely robust, and even so it would face serious risk over time. For this reason lunar cave systems have often been proposed as alternatives to building the whole thing ourselves.
NASA taps CMU to develop robots to help turn pits on the Moon into potential habitats
Nestled scores of meters below the surface’s fluffy (yet sharp) regolith and the harder layers beneath that, these tunnels are insulated from temperature fluctuations, protected from radiation and inaccessible to wandering space rocks. It’s just that until recently, such tunnels have been largely theoretical. We deduce their existence from pits like this one, which suggest a network of hollow tubes, but this paper is the first time anyone appears to have directly observed one (even if it was sort of by accident).
Diagram showing direction of radar reflections and likely shapes of cave interior.Image Credits: Carrer et al.
What this means for the space industry is that theory has become reality: The possibility of designing a habitation within a lunar tunnel is a reasonable proposition and one that is likely to be pursued with vigor and much funding after this. The location is even good: Mare Tranquillitatis, where Apollo 11 touched down (some distance away; it’s a big mare).
After the historic success of Intuitive Machines last year with the first-ever private moon landing, don’t be surprised if we see private and public efforts to reach and explore this cave specifically, or to establish the presence of others.
Not only that, but plans for lunar habitats and long-term presence are being drafted right now — and all of these plans must now be reevaluated in the context of a known and accessible moon cave. If NASA is going to spend $20 billion on something it hopes will last decades, it’ll want to get it right the first time — and this could be the next big notion in space.
As such, you can expect a wave of research and startups aimed at, if not outright troglodytic, at least cave-adjacent lunar tech. Lunar additive and robot manufacturers will show how they can shore up walls; solar providers will demonstrate drilled conduits or long-distance power transmission (more practical on the moon than here, actually). Robotic exploration and excavation will get a second wind. The lunar rover contenders will add bigger headlights.
Though the discovery of the moon cave is a shot in the arm for serious lunar habitat thinkers, builders and funders, it’s far from a sure thing. The cave is at the bottom of a pit, after all, meaning access will require excavation or infrastructure. Its dimensions and conditions aren’t known; it could easily be too small, crowded with rubble or otherwise impractical for habitation. But no one really expected the first tunnel we found to be a home run; its very existence changes the game, acting as a starting gun for the race to find the best moon cave in which to settle future lunar colonists.
Image Credits: Healx / CEO and co-founder Tim Guilliams
Healx, a U.K. startup using AI to discover new drugs for rare diseases, has raised $47 million in a Series C round of funding co-led by Europe’s Atomico and Silicon Valley-based R42.
The company also said that it has received regulatory clearance to start Phase 2 clinical trials for a new drug in the U.S. later this year.
Healx pitches itself as a “drug discovery pipeline” that identifies hidden connections between existing chemical compounds (e.g., drugs and pharmaceuticals at various stages of the development cycle) and rare diseases. It does this by collating a myriad of public and proprietary data sources, like biomedical literature, disease and biochemical datasets, clinical trials and patents to create a “biomedical knowledge graph” of rare diseases.
“Our AI platform and expert teams enable us to identify and discover novel disease biology and match it with suitable small molecules,” Healx co-founder and CEO Tim Guilliams told TechCrunch. “Our digital biology team focuses on understanding complex biological disease signatures, while our digital chemistry team uses AI to match these signatures with chemicals, small molecules and drugs, using computational techniques such as virtual screens and generative chemistry to do so.”
Typically, pharmaceutical companies adopt what is often referred to as a “one disease, one target, one drug” approach to drug discovery. This means they start by focusing on a single molecular target that’s responsible for a disease, and then set about designing a drug to target that molecule. The problem, however, is that this is hugely time-consuming and has high failure rates, which is where Healx promises to help by analyzing millions of drug and disease data points in tandem to find hitherto undetected connections that could present new treatments using known compounds.
Founded out of Cambridge in 2014, Healx is the brainchild of Guilliams and the company’s chairman, David Brown. Guilliams has a PhD in biophysics and neuroscience from the University of Cambridge, while Brown was formerly global head of drug discovery at Roche, though he is perhaps better known as the co-inventor of famed erectile dysfunction drug, Viagra.
Healx co-founders Tim Guilliams and David Brown.Image Credits: Healx
Drug design
With the fresh $47 million, Healx is gearing up to launch a Phase 2 clinical trial for a novel treatment it developed for neurofibromatosis Type 1 (NF1), a genetic disorder that causes tumor growth on human nerves. NF1 is considered one of the more common genetic disorders, affecting up to 1 in 3,000 individuals. While the tumors are usually benign, they can cause other symptoms related to the skin, eyes and nervous system, and they can also eventually become cancerous.
Healx had previously secured investigational new drug (IND) approval from the U.S. Food and Drug Administration (FDA) in 2021 for fragile X syndrome, and the company is also in the early AI-discovery or preclinical stages for a number of other conditions spanning oncology (cancer), renal (kidney), and neurology (nervous system).
Healx’s treatment pipeline.Image Credits: Healx
The Phase 2 trials of Healx’s NF1 lead candidate, dubbed HLX-1502 (a tablet taken orally), are expected to start by the end of 2024 and will focus on young adults with NF1 and inoperable plexiform neurofibroma, a specific type of tumor associated with NF1.
The company arrived at HLX-1502 using an existing drug developed for something else entirely that had never launched in the U.S. The company says it is reformulating that drug, and because it already had existing safety data attached to it due to previous testing in healthy patients, it gives the company a greater degree of confidence that it will have fewer side effects.
This is important, given that Healx is focusing on a disease that often involves benign tumors in otherwise healthy patients.
“There’s a lot of cancer treatments out there, where, basically, they kill everything — the healthy cells, the unhealthy cells — and there are a lot of side effects,” Guilliams said. “So for the patient community and the families, for those who have the benign form, it can live 40 to 50 years. It doesn’t make sense to put them on chronic cancer drugs for decades with side effects that are very strong.”
Healx’s technology looks at all manner of existing drugs in development, as well as those that have been shelved for various reasons but have enough data to make them useful for others to build on.
“There’s hundreds of thousands of compounds and small molecules,” Guilliams said. “But once you start seeing patterns between different data types and predictions, you can try to select the drug that you believe is the most de-risked. AI helps us to find the best possible drug for a particular disease biology. So here, we wanted something that went specifically to the nerve, where we believed there would be fewer side effects.”
Healx in action.Image Credits: Healx
All this leads to one obvious question: If Healx has developed its leading NF1 candidate by building on work previously carried out by others, does it actually own this candidate? According to Guilliams, Healx filed and owns the patents for HLX-1502 specifically for the use of treating NF1 and “other nerve-related diseases,” and the FDA has also given it new drug product exclusivity in the U.S. market. This is usually given out for new chemical entities (NCEs), which are defined as drugs that “are innovative” or where there have been “significant changes in already approved drug products, such as a new use.”
“Comprehensive commercial and IP analysis is always conducted when selecting our lead drug candidate taken to the clinic,” Guilliams said.
However, Healx didn’t reveal the actual origins of the original compound or the safety data.
There’s money in drugs
Before this round, Healx had raised around $68 million, the bulk of which came via a $56 million Series B round in 2019. The company also recently announced an investment from one of its long-term partners, the Children’s Tumor Foundation (CTF). The CTF will provide “milestone-driven payments” to Healx as it works toward and through its clinical trials.
It’s worth noting that although the CTF is a nonprofit, this financial arrangement isn’t a grant, though Healx hasn’t revealed the terms of the investment.
What’s particularly notable is that Healx’s Series C round is $9 million less than its Series B round five years ago, which, while not entirely unheard of, is a little unusual in the startup-funding realm. Indeed, when asked what the most capital-intensive part of Healx’s business was, Guilliams said that it’s carrying out clinical trials — a phase the company is only now embarking on with its NF1 candidate.
A source familiar with the matter told TechCrunch that Healx initially tried to raise significantly more than what it had hoped for its Series C round, and the company also cut its headcount by roughly 45% last year via a series of voluntary and compulsory redundancies — around 70 staff.
Healx confirmed as much to TechCrunch, with Guilliams pointing to the economic downturn and macroeconomic factors being less than kind to the startup fraternity these past few years.
“It is important to note that when the markets crashed, we were indeed in the process of raising a larger round,” Guilliams said. “We pulled that offering off the table at that time, and we then adapted effectively to the new market conditions by streamlining operations, refocusing on our core strengths, and becoming more cost-effective and lean. This allowed us to maintain our momentum and continue making significant progress, which we are proud to announce now. Our advancements in generative AI have further enhanced our capital efficiency, enabling us to achieve more with fewer resources. We then opened up the new financing round with a better perspective on what was needed to accomplish our goals.”
Thus, Guilliams is presenting the latest incarnation of its Series C fundraise as being oversubscribed, noting that it has now raised more than it targeted.
“Given the challenging market conditions for deep tech and biotech funding over the past two years, our ability to secure this investment underscores the confidence our investors have in our approach and technology,” he added.
Healx’s raise comes amid a swathe of investments into AI-driven drug discovery platforms. These include Sam Altman-backed Formation Bio, which recently raised $372 million, as well as a new startup called Xaira that launched with $1 billion in funding — both significantly more than what Healx has raised in total.
Guilliams stressed that much of the foundation work required to get its first drug to market has been done already, so it doesn’t necessarily require as much capital this time around. He also noted that his company’s valuation has gone up since its last funding round, but declined to disclose specifics.
“What’s important is the quality of your programs — picking those that have the most potential,” Guilliams said. “The market size for NF1 is $16 billion [according to Healx’s own internal commercial analysis], and if [its candidate] works on NF1, it’s likely to work on other nerve-related diseases — we’re testing those pre-clinically already. In rare diseases, the clinical trials are smaller, so we don’t need a ton of money, which, in the current markets, is a good place to be.”
Guilliams also confirmed the company did reduce redundancies in staff roles, though declined to go into specifics around numbers or circumstances beyond the fact that it included “both voluntary and compulsory measures.”
“Despite the challenge of letting go of exceptional people, this process ultimately made us a stronger and more focused team,” he said. “This restructuring was part of our strategy to streamline operations and enhance our focus on key therapeutic areas and technological advancements.”
Market entry
Today, Healx counts around 55 employees, roughly half of whom “code,” be that in the realm of machine learning or computational biology. The other half are the clinical and pharmacology team, who “work hand-in-hand with the techies,” Guilliams said.
While the bulk of its team is based in Cambridge (U.K.), the company is now looking to expand on the ground in the U.S., where its clinical trials are set to start by December. After that, it’s likely to be several more years before anything comes close to hitting the market due to the stringent regulatory requirements for new drugs.
“Conservatively, I would say we would expect clinical results the first half of 2026, and then depending on those results and interactions with the regulators, we may need to do a second clinical trial — and then potentially register the drug,” Guilliams said.
Aside from lead backers Atomico and R42, Healx’s Series C round included participation from Ayana Capital, Balderton, Btov, Global Brain, Jonathan Milner, O2h and VU Venture Partners.
Image Credits: Healx / CEO and co-founder Tim Guilliams
Healx, a U.K. startup using AI to discover new drugs for rare diseases, has raised $47 million in a Series C round of funding co-led by Europe’s Atomico and Silicon Valley-based R42.
The company also said that it has received regulatory clearance to start Phase 2 clinical trials for a new drug in the U.S. later this year.
Healx pitches itself as a “drug discovery pipeline” that identifies hidden connections between existing chemical compounds (e.g. drugs and pharmaceuticals at various stages of the development cycle) and rare diseases. It does this by collating myriad public and proprietary data sources, like biomedical literature, disease and biochemical datasets, clinical trials and patents to create a “biomedical knowledge graph” of rare diseases.
“Our AI platform and expert teams enable us to identify and discover novel disease biology and match it with suitable small molecules,” Healx co-founder and CEO Tim Guilliams told TechCrunch. “Our digital biology team focuses on understanding complex biological disease signatures, while our digital chemistry team uses AI to match these signatures with chemicals, small molecules and drugs, using computational techniques such as virtual screens and generative chemistry to do so.”
Typically, pharmaceutical companies adopt what is often referred to as a “one disease, one target, one drug” approach to drug discovery. This means they start by focusing on a single molecular target that’s responsible for a disease, and then set about designing a drug to target that molecule. The problem, however, is that this is hugely time-consuming and has high failure rates, which is where Healx promises to help by analyzing millions of drug and disease data points in tandem to find hitherto undetected connections that could present new treatments using known compounds.
Founded out of Cambridge in 2014, Healx is the brainchild of Guilliams and the company’s chairman, David Brown. Guilliams has a PhD in biophysics and neuroscience from the University of Cambridge, while Brown was formerly global head of drug discovery at Roche, though he is perhaps better known as the co-inventor of famed erectile dysfunction drug, Viagra.
Healx co-founders Tim Guilliams and David Brown.Image Credits: Healx
Drug design
With the fresh $47 million, Healx is gearing up to launch a Phase 2 clinical trial for a novel treatment it developed for Neurofibromatosis Type 1 (NF1), a genetic disorder that causes tumor growth on human nerves. Though a rare disease, NF1 is considered one of the more common genetic disorders, affecting up to 1 in 3,000 individuals. While the tumors are usually benign, they can cause other symptoms related to the skin, eyes and nervous system, and they can also eventually become cancerous.
Healx had previously secured investigational new drug (IND) approval from the U.S. Food and Drug Administration (FDA) in 2021 for Fragile X syndrome, and the company is also in the early AI-discovery or preclinical stages for a number of other conditions spanning oncology (cancer), renal (kidney), and neurology (nervous system).
Healx’s treatment pipeline.Image Credits: Healx
The Phase 2 trials of Healx’s NF1 lead candidate, dubbed HLX-1502 (a tablet taken orally), are expected to start by the end of 2024, and will focus on young adults with NF1 and inoperable plexiform neurofibroma, a specific type of tumor associated with NF1.
The company arrived at HLX-1502 using an existing drug developed for something else entirely that had never launched in the U.S. The company says it is reformulating that drug, and because it already had existing safety data attached to it due to previous testing in healthy patients, it gives the company a greater degree of confidence that it will have fewer side-effects.
This is important, given that Healx is focusing on a disease that often involves benign tumors in otherwise healthy patients.
“There’s a lot of cancer treatments out there, where, basically, they kill everything — the healthy cells, the unhealthy cells — and there are a lot of side effects,” Guilliams said. “So for the patient community and the families, for those who have the benign form, it can live 40 to 50 years. It doesn’t make sense to put them on chronic cancer drugs for decades with side-effects that are very strong.”
Healx’s technology looks at all manner of existing drugs in development, as well as those that have been shelved for various reasons but have enough data to make them useful for others to build on.
“There’s hundreds of thousands of compounds and small molecules,” Guilliams said. “But once you start seeing patterns between different data types and predictions, you can try to select the drug that you believe is the most de-risked. AI helps us to find the best possible drug for a particular disease biology. So here, we wanted something that went specifically to the nerve, where we believed there would be fewer side-effects.”
Healx in action.Image Credits: Healx
All this all leads to one obvious question: If Healx has developed its leading NF1 candidate by building on work previously carried out by others, does it actually own this candidate? According to Guilliams, Healx filed and owns the patents for HLX-1502 specifically for the use of treating NF1 and “other nerve-related diseases,” and the FDA has also given it new drug product exclusivity in the U.S. market. This is usually given out for new chemical entities (NCEs), which is defined as drugs that “are innovative” or where there have been “significant changes in already approved drug products, such as a new use.”
“Comprehensive commercial and IP analysis is always conducted when selecting our lead drug candidate taken to the clinic,” Guilliams said.
However, Healx didn’t reveal the actual origins of the original compound or the safety data.
There’s money in drugs
Before this round, Healx had raised around $68 million, the bulk of which came via a $56 million Series B round in 2019. The company also recently announced an investment from one of its long-term partners, the Children’s Tumor Foundation (CTF). The CTF will provide “milestone-driven payments” to Healx as it works towards and through its clinical trials.
It’s worth noting that although the CTF is a nonprofit, this financial arrangement isn’t a grant, though Healx hasn’t revealed the terms of the investment.
What’s particularly notable is that Healx’s Series C round is $9 million less than its Series B round five years ago, which, while not entirely unheard of, is a little unusual in the startup funding realm. Indeed, when asked what the most capital-intensive part of Healx’s business was, Guilliams said that it’s carrying out clinical trials — a phase the company is only now embarking on with its NF1 candidate.
A source familiar with the matter told TechCrunch that Healx initially tried to raise significantly more than what it had hoped for its Series C round, and the company also cut its headcount by roughly 45% last year via a series of voluntary and compulsory redundancies — around 70 staff.
Healx confirmed as much to TechCrunch, with Guilliams pointing to the economic downturn and macroeconomic factors being less than kind to the startup fraternity these past few years.
“It is important to note that when the markets crashed, we were indeed in the process of raising a larger round,” Guilliams said. “We pulled that offering off the table at that time, and we then adapted effectively to the new market conditions by streamlining operations, refocusing on our core strengths, and becoming more cost-effective and lean. This allowed us to maintain our momentum and continue making significant progress, which we are proud to announce now. Our advancements in generative AI have further enhanced our capital efficiency, enabling us to achieve more with fewer resources. We then opened up the new financing round with a better perspective on what was needed to accomplish our goals.”
Thus, Guilliams is presenting the latest incarnation of its Series C fundraise as being oversubscribed, noting that it has now raised more than it targeted.
“Given the challenging market conditions for deep tech and biotech funding over the past two years, our ability to secure this investment underscores the confidence our investors have in our approach and technology,” he added.
Healx’s raise comes amid a swathe of investments into AI-driven drug discovery platforms. These include Sam Altman-backed Formation Bio, which recently raised $372 million, as well as a new startup called Xaira that launched with $1 billion in funding — both significantly more than what Healx has raised in total.
Guilliams stressed that much of the foundation work required to get its first drug to market has been done already, so it doesn’t necessarily require as much capital this time around. He also noted that his company’s valuation has gone up since its last funding round, but declined to disclose specifics.
“What’s important is the quality of your programs — picking those that have the most potential,” Guilliams said. “The market size for NF1 is $16 billion [according to Healx’s own internal commercial analysis], and if [its candidate] works on NF1, it’s likely to work on other nerve-related diseases — we’re testing those pre-clinically already. In rare diseases, the clinical trials are smaller, so we don’t need a ton of money, which, in the current markets, is a good place to be.”
Guilliams also confirmed the company did reduce redundancies in staff roles, though declined to go into specifics around numbers or circumstances beyond the fact that it included “both voluntary and compulsory measures.”
“Despite the challenge of letting go of exceptional people, this process ultimately made us a stronger and more focused team,” he said. “This restructuring was part of our strategy to streamline operations and enhance our focus on key therapeutic areas and technological advancements.”
Market entry
Today, Healx counts around 55 employees, roughly half of whom “code,” be that in the realm of machine learning or computational biology. The other half are the clinical and pharmacology team, who “work hand-in-hand with the techies,” Guilliams said.
While the bulk of its team is based in Cambridge (U.K.), the company is now looking to expand on the ground in the U.S., where its clinical trials are set to start by December. After that, it’s likely to be several more years before anything comes close to hitting the market due to the stringent regulatory requirements for new drugs.
“Conservatively, I would say we would expect clinical results the first half of 2026, and then depending on those results and interactions with the regulators, we may need to do a second clinical trial — and then potentially register the drug,” Guilliams said.
Aside from lead backers Atomico and R42, Healx’s Series C round included participation from Ayana Capital, Balderton, Btov, Global Brain, Jonathan Milner, O2h and VU Venture Partners.
Scientists have identified what they believe is an accessible tunnel or cave on the moon. Such a feature, if real, may well define years of development by startups, governments and space companies aiming to create a lasting lunar colony.
Italian astronomers led by Leonardo Carrer and Lorenzo Bruzzone, working with Capella Space and JHUAPL, analyzed data collected by the Lunar Reconnaissance Orbiter in 2010. Specifically, its radar imagery of the surface, which in one place the team proposes was reflected in such a way that implies the presence of “a subsurface cave tens of meters long” — but potentially much larger.
It’s still at the bottom of a pit around 100 meters deep, but we’ll figure that part out later. What matters here is that we appear to have discovered a highly stable (otherwise it would have collapsed some time in the last few million years) moon cave that could serve as a base for lunar operations. Their findings are published in Nature Astronomy.
You may well ask: Why would we bother hiding in some ancient lava tunnel? Don’t we want to just make a surface habitat?
That’s certainly what we envision from science fiction, yes, but the reality is that the moon’s surface is quite an inhospitable place. With no ionosphere, it has no protection from solar or cosmic radiation, and its pocked surface shows how frequently it is bombarded by meteorites large and small, which are not slowed or burned up in an atmosphere. The temperature also varies from deep-space low to dangerously hot.
These and other factors mean that any surface dwelling would need to be extremely robust, and even so it would face serious risk over time. For this reason lunar cave systems have often been proposed as alternatives to building the whole thing ourselves.
NASA taps CMU to develop robots to help turn pits on the Moon into potential habitats
Nestled scores of meters below the surface’s fluffy (yet sharp) regolith and the harder layers beneath that, these tunnels are insulated from temperature fluctuations, protected from radiation and inaccessible to wandering space rocks. It’s just that until recently, such tunnels have been largely theoretical. We deduce their existence from pits like this one, which suggest a network of hollow tubes, but this paper is the first time anyone appears to have directly observed one (even if it was sort of by accident).
Diagram showing direction of radar reflections and likely shapes of cave interior.Image Credits: Carrer et al.
What this means for the space industry is that theory has become reality: The possibility of designing a habitation within a lunar tunnel is a reasonable proposition and one that is likely to be pursued with vigor and much funding after this. The location is even good: Mare Tranquillitatis, where Apollo 11 touched down (some distance away; it’s a big mare).
After the historic success of Intuitive Machines last year with the first-ever private moon landing, don’t be surprised if we see private and public efforts to reach and explore this cave specifically, or to establish the presence of others.
Not only that, but plans for lunar habitats and long-term presence are being drafted right now — and all of these plans must now be reevaluated in the context of a known and accessible moon cave. If NASA is going to spend $20 billion on something it hopes will last decades, it’ll want to get it right the first time — and this could be the next big notion in space.
As such, you can expect a wave of research and startups aimed at, if not outright troglodytic, at least cave-adjacent lunar tech. Lunar additive and robot manufacturers will show how they can shore up walls; solar providers will demonstrate drilled conduits or long-distance power transmission (more practical on the moon than here, actually). Robotic exploration and excavation will get a second wind. The lunar rover contenders will add bigger headlights.
Though the discovery of the moon cave is a shot in the arm for serious lunar habitat thinkers, builders and funders, it’s far from a sure thing. The cave is at the bottom of a pit, after all, meaning access will require excavation or infrastructure. Its dimensions and conditions aren’t known; it could easily be too small, crowded with rubble or otherwise impractical for habitation. But no one really expected the first tunnel we found to be a home run; its very existence changes the game, acting as a starting gun for the race to find the best moon cave in which to settle future lunar colonists.
Twitch is officially rolling out its new TikTok-like discovery feed to all users this week, the company announced on Tuesday. The new feed, which is launching as a tab within the Twitch iOS and Android apps, allows viewers to scroll through bite-sized bits of content to discover new streamers. The official launch follows Twitch’s earlier tests of the feed that started in August 2023.
With this launch, Twitch now joins numerous other popular apps that have launched their own short-form video feeds following TikTok’s rise in popularity, including Instagram, YouTube, Snapchat, Uber Eats and LinkedIn to name a few.
The discovery feed is finally here: the quickest way to find content you love on Twitch. Preview dozens of live streams & clips without pre-roll ads to find something to watch in minutes.
Update your Twitch mobile app to see the new Feed tab, rolling out to everyone this… pic.twitter.com/bI704lgYAB
— Twitch (@Twitch) April 30, 2024
There are two tabs within the feed: a live feed and a clips feed. The “live” feed displays livestreams from both streamers users already follow and streamers that Twitch thinks they would like. The “clips” feed showcases top moments from streamers across Twitch, allowing viewers to discover content from streamers even when they’re not live.
While people typically go to Twitch when they have enough time to watch a livestream, the new clips feed is designed for moments when users only have a few minutes to spare.
To better personalize what you see in the feed, you can click the thumbs-up or thumbs-down button on a livestream or a clip to signal what type of content you want to see more or less of.
The new feed’s arrival comes as Twitch streamers have long called on the company to introduce new features to help them reach new audiences. Since Twitch isn’t a fully algorithmic platform like TikTok, Twitch streamers haven’t had many ways to attract new viewers outside of promoting themselves on other platforms or teaming up with fellow streamers. With this new discovery feed, small and medium streamers have the potential to reach new audiences.
Twitch says that all livestreams and clips are automatically eligible to appear in the feed if they meet the platform’s content guidelines. While streamers can’t directly upload content to the discovery feed, Twitch encourages them to create and feature clips of their content for a chance to be included in the feed.
What’s next for Twitch? A big app redesign and more social sharing
