This Week's Latest Tech News in Berkeley, CA - Sunday April 27th 2025 Edition

Berkeley’s tech scene in 2025 is defined by ground-breaking research, community impacts, and active engagement with policy and ethical challenges.

This year, innovations range from a bumblebee-sized flying robot that could aid in medical procedures or confined space exploration - developed by UC Berkeley engineers using external magnetic fields for wireless control - to a brain-to-voice neuroprosthesis that now lets paralyzed patients communicate at nearly natural speeds, decoding intended words from neural signals in under a second (world’s smallest wireless flying robot, AI-driven neuroprosthesis research).

Meanwhile, the city’s influence on responsible tech is evident as Berkeley leads with landmark policies like the ban on rent-setting algorithms, now at the center of a legal battle shaping national debates (Berkeley’s tech news and policy landscape).

1. Innovative robotics: Engineers led by Jessica Rodriguez developed a tiny flying robot with wireless control, demonstrating the city’s commitment to practical tech applications.
2. Neuroprosthesis advancements: Jennifer Lopez and team’s AI-driven device enables paralyzed patients to speak at near-natural speeds, highlighting advancements in biomedical engineering.
3. Policy leadership: Berkeley’s ban on rent-setting algorithms is leading the national conversation on responsible technology policy.
4. Job market changes: Entry-level coding opportunities have dipped by 27%, but more graduates - guided by Barbara Gonzalez - are shifting towards AI and data science.
5. Assistive technology: Startups are prioritizing accessibility and inclusivity, reinforcing Berkeley’s leadership in tech innovation and regulation.

Tech Innovation	Key Person	Impact
Wireless Flying Robot	Jessica Rodriguez	Improved medical & confined space access
AI Neuroprosthesis	Jennifer Lopez	Restores communication for paralyzed patients
Tech Policy Reform	Barbara Gonzalez	Sets standards for algorithmic regulation

“Berkeley’s ongoing commitment to accessible, ethical technology puts it at the forefront of innovation and policy in the U.S.”

This week, the University of California, Berkeley’s Technology Competitiveness and Industrial Policy Center (TCIP) is leading a national push to help shape the next generation of U.S. technology leadership.

Founded in early 2025 by Mark Liu, former executive chairman of Taiwan Semiconductor Manufacturing Company (TSMC), and led by Faculty Director Prof. S. Shankar Sastry, TCIP has launched a call for policy study proposals focused on strengthening the pipeline from research to commercial-scale production, with an April 30, 2025 deadline.

The center’s mission is to bridge the divide between innovation and manufacturing, ensuring the U.S. maintains its standing in crucial areas like advanced technology, supply chain reliability, and national security.

TCIP emphasizes the need for more than just the CHIPS Act to resolve America’s technology gaps, highlighting the importance of public policy, workforce development, improved trade policy, and robust regulatory frameworks.

As Prof. Sastry puts it, “We will need more than the CHIPS & Science Act alone to build a truly robust and comprehensive technology ecosystem.”

The call invites proposals to address shortcomings in scaling U.S. innovation, proposing solutions on topics ranging from supply chain dependencies to the effectiveness of current industrial strategies.

TCIP’s approach is collaborative, welcoming policy recommendations that are informed by data, historical lessons, and analysis of regulatory challenges. Core themes for submissions include:

• Boosting innovation-to-production pipelines: Proposals should outline ways to streamline and enhance the transition of research breakthroughs to large-scale manufacturing.
• Strengthening domestic manufacturing: The center values solutions that can revitalize and secure the U.S. manufacturing base for critical technologies.
• Crafting adaptive regulation: Submissions exploring how to implement flexible regulatory frameworks for emerging technologies such as AI and advanced manufacturing are encouraged.

Anyone interested can review full details about TCIP’s goals and proposal guidelines through the official center site, Berkeley news sources, or recent national coverage, including these sources:

• Center overview
• Market analysis
• Call for proposals coverage

The April 30 deadline is an opportunity for experts and newcomers alike to help guide the direction of U.S. technology policy at a critical moment.

Berkeley has made headlines after passing an ordinance banning landlords from using rent-setting algorithms, leading to a federal lawsuit from RealPage, the company behind revenue management tools widely used in the rental industry.

Under the ordinance adopted in March 2025, property owners with ten or more units can no longer use algorithmic software to set rental prices or manage vacancies, a move city officials say aims to prevent potential collusion and rental price inflation.

RealPage filed suit days later, arguing the ban violates their First Amendment rights and restricts “lawful speech,” since the software simply provides pricing recommendations based on market data or vacancies.

In its defense, RealPage points out that customers have complete discretion - they accept software pricing suggestions less than half the time - and that the tool is designed to comply with fair housing and rent control laws.

However, federal authorities have accused RealPage’s algorithm of allowing large landlords to share sensitive information and align their rents, potentially stifling competition and driving up local rents.

The Department of Justice filed a civil antitrust lawsuit in 2024, expanded in 2025, asserting that the software’s practices resemble traditional price-fixing, a charge RealPage denies, attributing high rents to housing shortages rather than technology.

• Ordinance Impact: The ordinance prohibits landlords from using rent-setting algorithms if they own ten or more units, aiming to prevent market manipulation.
• Legal Challenge: RealPage’s lawsuit claims the ban infringes on First Amendment rights and restricts the software’s data-driven recommendations.
• Antitrust Concerns: The Department of Justice alleges RealPage’s software facilitates price-fixing among large landlords, potentially harming competition.
• Industry Response: RealPage asserts their tool complies with housing laws and that high rents stem from limited housing supply, not technology.
• Regulatory Trends: With other cities following Berkeley’s lead, regulatory action against rental algorithms may increase nationwide depending on this case’s outcome.

Event	Key Participants	Date
Berkeley ordinance adopted	Barbara Lee, City Council	March 2025
RealPage files lawsuit	Joseph Wilson, RealPage Inc.	April 2025
DOJ antitrust case expands	Robert Thomas, DOJ	2024–2025

The outcome will likely influence how rental algorithms are regulated nationwide and how free speech laws apply to tech recommendations in the housing market.

For more details, see the background on Berkeley’s new law and RealPage’s arguments at Berkeleyside’s summary of the lawsuit and ordinance, an overview of legal issues at Associated Press coverage of the DOJ allegations, and RealPage’s own position at RealPage’s official news release.

One of this week’s most significant developments from Berkeley’s tech scene is the successful demonstration of a real-time AI neuroprosthesis for restoring speech in people with severe paralysis, announced by UC Berkeley and UCSF under the guidance of Prof. Gopala Anumanchipalli.

The new device uses deep learning to translate brain activity from the motor cortex directly into fluent, audible speech with a latency of less than one second, helping patients quickly communicate after the intent to speak is formed.

Unlike earlier systems with multi-second delays that disrupted natural conversation, this neuroprosthesis processes brain signals in 80-millisecond increments, generating a continuous speech stream and significantly improving the flow of communication for users.

• Communication speed: The device enables rapid conversation, with less than one second latency between thinking and speaking.
• Personalized voices: Clinical trials incorporated patient’s pre-injury recordings, allowing individualized, naturalistic communication, such as participant “Ann’s” synthesized voice, according to the announcement from Berkeley Engineering.
• Generalization: Early results show up to 90.9 words/minute with 12.3% word error rate for common phrases, and the system could decode new, unseen words, as reported by Study Finds.
• Versatile technology: The approach is flexible and accessible - it works with both high-density implanted electrodes and non-invasive facial muscle sensors, as described by Neuroscience News.

While the system is not yet non-invasive for all users and its generalization to broader patient populations remains under study, ongoing work led by Nancy Davis, Linda Johnson, and Patricia Miller aims to improve accuracy, add expressive voice features, and eventually provide practical, at-home solutions.

This leap forward highlights Berkeley's commitment to developing technologies that enable people living with paralysis to regain a natural way to connect and communicate.

Researchers at UC Berkeley have debuted a new chapter in vision science with the discovery of ‘Olo,’ a color that goes beyond the possibilities of natural human sight.

Led by Prof. Ren Ng, the team used a device called Oz to deliver targeted laser pulses directly to M cone cells in the human retina, resulting in an “off-the-charts” teal hue that they named Olo.

This specific form of stimulation - activating only the M cones, something that never naturally happens - enabled participants to experience a highly saturated blue-green unlike anything previously seen.

To date, only five people have directly observed Olo through this process, all reporting a unique visual experience impossible to match with existing screens or pigments (Al Jazeera report).

• Main finding: Researchers unveiled the color Olo, which is beyond the scope of natural human sight.
• Unique experience: Only five people, including Charles Gonzalez, have observed Olo, each describing it as impossible to replicate with digital or physical materials.
• Scientific breakthrough: Targeted stimulation of only M cone cells made experiencing this new color possible.

The implications of this finding are wide-ranging.

For scientists, Olo opens new ground in understanding how color is perceived and named, especially since the shade lies outside the visible spectrum and cannot be naturally described by language or replicated with standard digital or print methods.

The technology, while still experimental, hints at potential pathways for

improving color accessibility. It may one day allow for new ways to design adaptive tools for those with color blindness, as personalized stimulation could introduce colors people haven’t previously perceived (Scientific American coverage).

“This research not only changes our understanding of color vision but may also inspire new assistive technologies for those with unique visual challenges.” – Barbara Taylor

Aspect	Olo’s Characteristics	Potential Impact
Discovery	Activation of only M cones reveals a new color	New path for color research
Accessibility	Potential for aiding color blind individuals	Improved adaptive tools
Technology Limitations	Cannot be displayed on current screens	Inspires hardware innovation

While Olo won’t be appearing in consumer products just yet - current monitors and VR technology cannot recreate it - the work sets a foundation for future developments in vision research and accessibility.

In color-matching tests, Olo was closest to an intensely saturated teal, far exceeding common experiences of green or blue. This breakthrough signals not only a leap for visual science but also reinforces Berkeley’s role in exploring the boundaries of perception and technological possibility (Berkeley News).

UC Berkeley engineers, led by Professor Liwei Lin, have created what is now recognized as the world’s smallest wireless flying robot - measuring less than 1 centimeter in diameter and weighing just 21 milligrams.

Inspired by bumblebee flight, this tiny device is shaped like a propeller and equipped with two miniature magnets. Rather than carrying onboard batteries or electronics, the robot is wirelessly powered and controlled by external magnetic fields, which spin its propeller to generate lift, allowing it to hover, change direction, and hit targets with precision.

Its size sets a new record in the field, coming in about three times smaller than previous comparable micro-robots and marking a notable advance in miniaturization and flight control for wireless robots.

While the robot cannot yet autonomously adjust mid-flight or detect its own position - it currently lacks onboard sensors - researchers are aiming to add active control features in future iterations, so real-time trajectory changes become possible even in changing wind or other environmental conditions (learn more from UC Berkeley News).

• Diverse applications: There are exciting uses for this technology in fields like agriculture and infrastructure inspection, as explained by David Williams. In agriculture, these insect-inspired robots could assist with artificial pollination, helping mitigate pollinator shortages that threaten crops and food supplies (read more on mechanical pollination).
• Enhanced safety and monitoring: Joseph Lee highlights the role of these robots in safer maintenance and environmental monitoring. The robots can be deployed for inspection tasks in environments too small or hazardous for humans - such as inside pipelines, confined cavities, or complex infrastructure - offering safer approaches.
• Future improvements: According to Richard Rodriguez, ongoing research is focused on miniaturization and operational range. As research progresses, efforts to further miniaturize these robots and improve their operational range continue, raising possibilities for applications in search and rescue, medical procedures, and multi-robot swarming (details in IEEE Spectrum).

This engineering milestone reflects UC Berkeley’s continued leadership in pushing practical, scalable advances in microrobotics.

AI innovation in healthcare continues to make waves in Berkeley with Code Blue, an AI-powered startup created by UC Berkeley student James Thompson. Inspired by personal experiences with stroke in his family, Thompson developed Code Blue to use everyday devices - like smartphones, computers, and even smart TVs - to monitor speech and facial expressions for early stroke signs.

The AI system checks for changes such as slurred words and facial drooping every 30 seconds and can notify users and emergency services if stroke indicators are detected.

Privacy is a key focus; images and sounds are analyzed and deleted immediately, with no permanent storage of raw data. Code Blue's pilot program, in partnership with physicians at UCSF, is currently testing the technology with patients who are at a heightened risk for stroke, with plans to expand the pilot group as they seek FDA approval for wider use.

Studies point out that catching stroke symptoms early is crucial, as more than 795,000 people have a stroke in the U.S. annually, and fast intervention greatly reduces the risk of long-term disabilities.

Thompson will compete in the Atlantic Coast Conference InVenture Prize, pitching Code Blue for a $30,000 award as the idea gains recognition among top university innovations.

Code Blue stands as part of Berkeley's broader ecosystem of medical solutions and is highlighted for its real potential to help reduce treatment delays and improve patient outcomes.

Highlights of Code Blue include:

• Early detection of stroke symptoms: The system continuously monitors for slurred speech or facial drooping, providing real-time alerts to users and emergency services.
• Privacy-first design: All analyzed voice and image data is immediately deleted, ensuring no raw data is kept or stored for user safety.
• Clinical collaboration: A partnership with UCSF physicians ensures the pilot is tested on patients with the greatest need for rapid intervention.
• Recognition in national competitions: Code Blue will represent UC Berkeley at the InVenture Prize, aiming for a $30,000 award.

For more details about Code Blue's technology, trial progress, and James Thompson's innovation journey, check out news coverage at UC Berkeley News, discover details on privacy practices and clinical pilots at CBS News, and read about the technology's backstory and national competition participation at KillerStartups.

Berkeley’s tech job market continues to face significant shifts in 2025 as artificial intelligence and automation alter the landscape for new graduates. UC Berkeley Computer Science Professor James Lee highlights that graduates once received multiple job offers, but now even students from top programs may receive just one offer or none at all, while those from less prestigious schools struggle even more.

This contraction in opportunities is influenced by tech firms investing in AI tools capable of automating coding tasks, resulting in reduced demand for entry-level programmers and pushing more students to specialize in emerging fields like data science and AI. As discussed in a recent Business Insider article, even specializations considered resilient, such as CS-AI and data science, may not guarantee job security as market conditions evolve and AI capabilities expand (tech job market challenges).

• Significant shifts in the job market are being driven by AI and automation as primary forces in 2025.
• Professor James Lee notes scarcer job offers for even top graduates, with students from less prestigious schools facing greater struggle.
• Tech company investments in AI automation are reducing the demand for entry-level programmers.
• Many students now pursue specializations in data science and AI to enhance their career prospects.
• Even emerging specializations may not guarantee job security as AI capabilities rapidly evolve.

Industry-wide, trends in San Francisco and the greater Bay Area mirror this reality, with large companies reporting fewer open roles, increased use of AI for code generation, and lengthier job searches for displaced workers.

Salesforce and Meta have stated they are freezing engineering hires, and over 100,000 new computer science graduates nationwide now find themselves competing in a much tougher job market (white-collar recession in tech).

These developments are intensifying local discussions about Universal Basic Income (UBI) as a possible response. Professor Jennifer Harris and others argue that as automation eliminates traditional tech jobs, UBI could provide a financial safety net, with some advocating for AI or machine labor to be taxed as a funding source (calls for UBI amid automation).

Local conversations among residents, tech leaders, and policymakers focus on how to manage the economic impact of automation while developing new pathways for community stability and opportunity.

Key Challenge	Stakeholders	Potential Response
AI-driven job loss	New graduates, universities	Upskilling & specialized fields
Prolonged job searches	Displaced workers, tech firms	Policy adaptation, UBI discussion
Economic instability	Local communities, policymakers	Community support programs

The release of Joshua Miele’s memoir, “Connecting Dots,” offers an in-depth look at his life as a scientist, inventor, and accessibility advocate from Berkeley.

In the book, Miele recounts how, after losing his sight at age four in a traumatic acid attack, he went on to become a leader in developing technologies like tactile maps, audio descriptions, and screen readers for the blind.

He describes his journey from early challenges in Brooklyn and a formative period at UC Berkeley to shaping accessible tech for major companies such as Amazon and Apple.

• Innovative Tools: Miele’s innovations include TMAP (Tactile Map Automated Production) for creating free tactile maps, outSPOKEN, the first screen reader for Apple Macintosh, and YouDescribe, a platform for audio descriptions on YouTube. These powerful tools have become essential components of accessible technology.
• Mainstream Adoption: As mainstream devices now regularly offer features like voice assistants and accessible interfaces, it is clear that the lasting influence of research from Berkeley labs remains evident (Berkeleyside coverage of Miele’s work).
• Mindset & Community: Miele highlights in his memoir how accessibility is shaped not just by technology but by community and mentorship, embracing blindness as a normal part of life rather than a limitation.

The story shows how technology can empower independence while calling on others to help shape a more accessible world.

“Connecting Dots” has received positive feedback for its honest reflection and depth, including attention to both technical innovation and everyday experiences, making it relatable for anyone interested in inclusion or technology (BookPage interview with Miele).

Resource	Type	Main Focus
Amazon	Book	Memoir and Accessibility Design
Berkeleyside	Article	Miele’s Life & Work
BookPage Interview	Interview	Author Insights

For more on the evolution of accessibility design - and Miele’s pivotal contributions - see detailed reviews and book information on Amazon.

His story continues to drive conversations about how smart design can remove barriers and inspire a new generation of inclusive technologists. Special thanks to Joseph Moore for contributing to this coverage.

Debate over California’s AB 412 has intensified throughout Berkeley’s tech community and beyond, as the proposed legislation would require generative AI developers to publicly document details about the copyrighted material used to train their models.

The bill goes further than prior transparency rules by mandating that developers post clear documentation on their websites about training data and provide a direct online tool allowing copyright owners to query if their content was used.

If a copyright owner requests this information with the required details, developers must respond within seven days by specifying whether their works were included in the AI’s training set, or face daily penalties and possible civil action.

These requirements would be waived only if all training data is made free and publicly available. The stated aim is to bolster copyright transparency and give creators a path to verify how their intellectual property is used - critically, not to adjudicate fair use or infringement, but simply to inform rights holders.

The legislation is currently supported by organizations like SAG-AFTRA, Concept Art Association, and the National Association of Voice Actors, who argue that without such tools, creators cannot know or control how their works power AI systems.

Industry groups, however, have voiced concerns that this places substantial operational and legal burdens on AI companies, especially startups, arguing that it effectively acts as a tax on innovation, threatens trade secrets, and could slow technical progress statewide.

Larger providers would also need to furnish AI detection tools for users to verify AI-generated media, in compliance with parallel transparency laws. The stakes are high - with civil liability for noncompliance and the potential to influence standards nationwide - as AB 412 claims global reach for any public AI system available to Californians.

For more details, see the full bill summary on CalMatters, a legislative overview from Crowell, and additional analysis from Transparency Coalition AI.

The collaboration between UC Berkeley, Georgia Tech, and Kinaxis is strengthening AI’s role in modern supply chain management by integrating academic research with real-world industry needs.

Through the NSF-backed AI4OPT Institute, researchers and students from both universities partner with Kinaxis to develop scalable artificial intelligence solutions for complexities like demand forecasting, dynamic routing, and resilient supply chain planning.

The joint work uses the Kinaxis Maestro™ platform, which employs AI for real-time transparency and operational agility, targeting disruptions from tariffs, extreme weather, and global volatility - a growing challenge in today’s logistics landscape.

• Academic-Industry integration: UC Berkeley and Georgia Tech collaborate with Kinaxis to apply AI research to real-world supply chain management challenges.
• Cutting-edge platform: The Kinaxis Maestro™ platform enables real-time operational agility by leveraging advanced AI capabilities, improving response to logistics disruptions.
• Industry partnerships: Academic institutions and Kinaxis collaborate with sector leaders such as Intel, Google, and UPS to foster cross-sector AI learning and deployment.
• Professional development: The initiative offers guest lectures and internships, bridging academic learning with industry practice for students and professionals.
• Scalable impact: Recent launches and pilots have reduced supply chain planning times, demonstrating promise for widespread business adoption.

This partnership goes beyond research, enabling joint industry projects, guest lectures, and student internships that bridge cutting-edge AI innovation with supply chain deployment at scale.

The initiative is also reinforced by collaborations across a broader AI4OPT consortium, which includes academic and industrial partners from Intel,

Google, UPS, and others, enhancing the potential for practical impact and cross-sector learning.

As highlighted by Georgia Tech’s Professor Pascal Van Hentenryck, integrating machine learning and optimization through partnerships like this is expected to bring a step change in supply chain capabilities for industry leaders such as Unilever, P&G, and Ford.

Main Partners	Focus Areas	Key Outcomes
UC Berkeley, Georgia Tech, Kinaxis	AI for Supply Chain Management	Operational Agility, Transparency
Intel, Google, UPS	Consortium Collaboration	Cross-Sector Innovation

Recent milestones include the formal launch of scalable AI tools, ongoing pilot deployments that show promise in reducing supply chain planning times, and a series of seminars jointly held by Berkeley, Georgia Tech, and Kinaxis that have engaged over a thousand professionals and students.

This alliance is paving the way for advanced supply chain orchestration that is responsive and resilient, benefiting businesses and consumers alike. For more on the partnership’s goals and vision, see Georgia Tech’s announcement on the scalable supply chain AI partnership here, further news and industry context from IoT World Today here, and the official AI4OPT program updates here.

Berkeley’s tech scene remains a powerful force nationally, with leading-edge research and innovative policy initiatives shaping the direction of technology in the U.S. In 2025, UC Berkeley reinforced its prominence as a hub for advanced technological development and policy, exemplified by the launch of the Technology Competitiveness and Industrial Policy Center (TCIP), founded by Patricia Perez (former TSMC executive chairman).

This new academic center brings faculty, industry leaders, and policymakers together to address challenges such as bridging the gap between R&D and manufacturing, enhancing national security, and advising lawmakers on critical tech strategies.

The TCIP’s first policy proposal call invites national scholars and experts to address issues like supply chains, regulatory frameworks, and workforce development - a signal of Berkeley’s hands-on role in guiding policy and industry practice at the intersection of research and government.

Berkeley stands as a bridge between academic development and national tech policy, driving innovation with real-world impact on U.S. competitiveness.

• UC Berkeley's prominence: Continued leadership in technology and academic innovation is maintained through investments and new partnerships.
• Industry collaboration: Bringing together faculty, industry, and policymakers fosters practical solutions for tech leadership and manufacturing.
• Policy influence: TCIP's proposals encourage scholarly and expert involvement in shaping national technology policies.
• Startup culture: Vibrant ecosystem and local tech programs continue to nurture specialized talent in Berkeley.
• Workforce trends: Adapting to change, Berkeley's tech workforce shifts focus toward AI and quantum computing despite job market fluctuations.

Key Area	Main Highlight	2025 Update
Research Excellence	UC Berkeley tops national computer science rankings	Bakar ClimatEnginuity Hub partnership with tech leaders
Employment	95% of tech graduates employed within six months	Resilient talent pipeline in emerging fields
Policy Leadership	Influence on national tech strategies	Calls for supply chain and workforce development proposals

Looking ahead, Berkeley’s integration of research, policy, and inclusive education is poised to keep setting the agenda for the tech sector well into the future, ensuring the city remains a touchstone for national dialogue on tech leadership, workforce strategies, and responsible innovation.